config,proxy: add new configuration logToStdout (#432)

The new logToStdout option controls what is logged to stdout. The
default has been changed to just the proxy logs, which contain swap and
http request logs.

There are four supported settings: none, proxy, upstream, both. The
"both" setting is the legacy setting where everything was spewed to
stdout.

.github/workflows/config-schema.yml

@@ -0,0 +1,41 @@
+name: Validate JSON Schema
+
+on:
+  pull_request:
+    paths:
+      - "config-schema.json"
+  push:
+    branches:
+      - main
+    paths:
+      - "config-schema.json"
+
+  workflow_dispatch:
+
+jobs:
+  validate-schema:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Validate JSON Schema
+        run: |
+          # Check if the file is valid JSON
+          if ! jq empty config-schema.json 2>/dev/null; then
+            echo "Error: config-schema.json is not valid JSON"
+            exit 1
+          fi
+
+          # Validate that it's a valid JSON Schema
+          # Check for required $schema field
+          if ! jq -e '."$schema"' config-schema.json > /dev/null; then
+            echo "Warning: config-schema.json should have a \$schema field"
+          fi
+
+          # Check that it has either properties or definitions
+          if ! jq -e '.properties or .definitions or ."$defs"' config-schema.json > /dev/null; then
+            echo "Warning: JSON Schema should contain properties, definitions, or \$defs"
+          fi
+
+          echo "✓ config-schema.json is valid"

CLAUDE.md

@@ -11,7 +11,7 @@ llama-swap is a light weight, transparent proxy server that provides automatic m
 
 ## Testing
 
-- `make test-dev` - Use this when making iterative changes. Runs `go test` and `staticcheck`. Fix any static checking errors.
+- `make test-dev` - Use this when making iterative changes. Runs `go test` and `staticcheck`. Fix any static checking errors. Use this only when changes are made to any code under the `proxy/` directory
 - `make test-all` - runs at the end before completing work. Includes long running concurrency tests.
 
 ## Workflow Tasks

Makefile

@@ -90,7 +90,7 @@ GOOS ?= $(shell go env GOOS 2>/dev/null || echo linux)
 GOARCH ?= $(shell go env GOARCH 2>/dev/null || echo amd64)
 wol-proxy: $(BUILD_DIR)
 	@echo "Building wol-proxy"
-	go build -o $(BUILD_DIR)/wol-proxy-$(GOOS)-$(GOARCH) cmd/wol-proxy/wol-proxy.go
+	go build -o $(BUILD_DIR)/wol-proxy-$(GOOS)-$(GOARCH)-$(shell date +%Y-%m-%d) cmd/wol-proxy/wol-proxy.go
 
 # Phony targets
 .PHONY: all clean ui mac linux windows simple-responder simple-responder-windows test test-all test-dev wol-proxy

README.md

@@ -1,4 +1,4 @@
-![llama-swap header image](header2.png)
+![llama-swap header image](docs/assets/hero3.webp)
 ![GitHub Downloads (all assets, all releases)](https://img.shields.io/github/downloads/mostlygeek/llama-swap/total)
 ![GitHub Actions Workflow Status](https://img.shields.io/github/actions/workflow/status/mostlygeek/llama-swap/go-ci.yml)
 ![GitHub Repo stars](https://img.shields.io/github/stars/mostlygeek/llama-swap)
@@ -13,7 +13,7 @@ Built in Go for performance and simplicity, llama-swap has zero dependencies and
 
 - ✅ Easy to deploy and configure: one binary, one configuration file. no external dependencies
 - ✅ On-demand model switching
-- ✅ Use any local OpenAI compatible server (llama.cpp, vllm, tabbyAPI, etc)
+- ✅ Use any local OpenAI compatible server (llama.cpp, vllm, tabbyAPI, etc.)
   - future proof, upgrade your inference servers at any time.
 - ✅ OpenAI API supported endpoints:
   - `v1/completions`
@@ -21,6 +21,8 @@ Built in Go for performance and simplicity, llama-swap has zero dependencies and
   - `v1/embeddings`
   - `v1/audio/speech` ([#36](https://github.com/mostlygeek/llama-swap/issues/36))
   - `v1/audio/transcriptions` ([docs](https://github.com/mostlygeek/llama-swap/issues/41#issuecomment-2722637867))
+- ✅ Anthropic API supported endpoints:
+  - `v1/messages`
 - ✅ llama-server (llama.cpp) supported endpoints
   - `v1/rerank`, `v1/reranking`, `/rerank`
   - `/infill` - for code infilling
@@ -42,7 +44,7 @@ Built in Go for performance and simplicity, llama-swap has zero dependencies and
 
 llama-swap includes a real time web interface for monitoring logs and controlling models:
 
-<img width="1360" height="963" alt="image" src="https://github.com/user-attachments/assets/adef4a8e-de0b-49db-885a-8f6dedae6799" />
+<img width="1164" height="745" alt="image" src="https://github.com/user-attachments/assets/bacf3f9d-819f-430b-9ed2-1bfaa8d54579" />
 
 The Activity Page shows recent requests:
 
@@ -60,7 +62,7 @@ llama-swap can be installed in multiple ways
 
 ### Docker Install ([download images](https://github.com/mostlygeek/llama-swap/pkgs/container/llama-swap))
 
-Nightly container images with llama-swap and llama-server are built for multiple platforms (cuda, vulkan, intel, etc).
+Nightly container images with llama-swap and llama-server are built for multiple platforms (cuda, vulkan, intel, etc.) including [non-root variants with improved security](docs/container-security.md).
 
 ```shell
 $ docker pull ghcr.io/mostlygeek/llama-swap:cuda
@@ -70,6 +72,14 @@ $ docker run -it --rm --runtime nvidia -p 9292:8080 \
  -v /path/to/models:/models \
  -v /path/to/custom/config.yaml:/app/config.yaml \
  ghcr.io/mostlygeek/llama-swap:cuda
+
+# configuration hot reload supported with a
+# directory volume mount
+$ docker run -it --rm --runtime nvidia -p 9292:8080 \
+ -v /path/to/models:/models \
+ -v /path/to/custom/config.yaml:/app/config.yaml \
+ -v /path/to/config:/config \
+ ghcr.io/mostlygeek/llama-swap:cuda -config /config/config.yaml -watch-config
 ```
 
 <details>
@@ -88,6 +98,9 @@ docker pull ghcr.io/mostlygeek/llama-swap:musa
 # tagged llama-swap, platform and llama-server version images
 docker pull ghcr.io/mostlygeek/llama-swap:v166-cuda-b6795
 
+# non-root cuda
+docker pull ghcr.io/mostlygeek/llama-swap:cuda-non-root
+
 ```
 
 </details>
@@ -190,23 +203,26 @@ As a safeguard, llama-swap also sets `X-Accel-Buffering: no` on SSE responses. H
 
 ## Monitoring Logs on the CLI
 
-```shell
+```sh
 # sends up to the last 10KB of logs
-curl http://host/logs'
+$ curl http://host/logs
 
 # streams combined logs
-curl -Ns 'http://host/logs/stream'
+curl -Ns http://host/logs/stream
 
-# just llama-swap's logs
-curl -Ns 'http://host/logs/stream/proxy'
+# stream llama-swap's proxy status logs
+curl -Ns http://host/logs/stream/proxy
 
-# just upstream's logs
-curl -Ns 'http://host/logs/stream/upstream'
+# stream logs from upstream processes that llama-swap loads
+curl -Ns http://host/logs/stream/upstream
+
+# stream logs only from a specific model
+curl -Ns http://host/logs/stream/{model_id}
 
 # stream and filter logs with linux pipes
 curl -Ns http://host/logs/stream | grep 'eval time'
 
-# skips history and just streams new log entries
+# appending ?no-history will disable sending buffered history first
 curl -Ns 'http://host/logs/stream?no-history'
 ```
 

ai-plans/2025-12-14-efficient-ring-buffer.md

@@ -0,0 +1,85 @@
+# Replace ring.Ring with Efficient Circular Byte Buffer
+
+## Overview
+
+Replace the inefficient `container/ring.Ring` implementation in `logMonitor.go` with a simple circular byte buffer that uses a single contiguous `[]byte` slice. This eliminates per-write allocations, improves cache locality, and correctly implements a 10KB buffer.
+
+## Current Issues
+
+1. `ring.New(10 * 1024)` creates 10,240 ring **elements**, not 10KB of storage
+2. Every `Write()` call allocates a new `[]byte` slice inside the lock
+3. `GetHistory()` iterates all 10,240 elements and appends repeatedly (geometric reallocs)
+4. Linked list structure has poor cache locality and pointer overhead
+
+## Design Requirements
+
+### New CircularBuffer Type
+
+Create a simple circular byte buffer with:
+- Single pre-allocated `[]byte` of fixed capacity (10KB)
+- `head` and `size` integers to track write position and data length
+- No per-write allocations
+
+### API Requirements
+
+The new buffer must support:
+1. **Write(p []byte)** - Append bytes, overwriting oldest data when full
+2. **GetHistory() []byte** - Return all buffered data in correct order (oldest to newest)
+
+### Implementation Details
+
+```go
+type circularBuffer struct {
+    data []byte  // pre-allocated capacity
+    head int     // next write position
+    size int     // current number of bytes stored (0 to cap)
+}
+```
+
+**Write logic:**
+- If `len(p) >= capacity`: just keep the last `capacity` bytes
+- Otherwise: write bytes at `head`, wrapping around if needed
+- Update `head` and `size` accordingly
+- Data is copied into the internal buffer (not stored by reference)
+
+**GetHistory logic:**
+- Calculate start position: `(head - size + cap) % cap`
+- If not wrapped: single slice copy
+- If wrapped: two copies (end of buffer + beginning)
+- Returns a **new slice** (copy), not a view into internal buffer
+
+### Immutability Guarantees (must preserve)
+
+Per existing tests:
+1. Modifying input `[]byte` after `Write()` must not affect stored data
+2. `GetHistory()` returns independent copy - modifications don't affect buffer
+
+## Files to Modify
+
+- `proxy/logMonitor.go` - Replace `buffer *ring.Ring` with new circular buffer
+
+## Testing Plan
+
+Existing tests in `logMonitor_test.go` should continue to pass:
+- `TestLogMonitor` - Basic write/read and subscriber notification
+- `TestWrite_ImmutableBuffer` - Verify writes don't affect returned history
+- `TestWrite_LogTimeFormat` - Timestamp formatting
+
+Add new tests:
+- Test buffer wrap-around behavior
+- Test large writes that exceed buffer capacity
+- Test exact capacity boundary conditions
+
+## Checklist
+
+- [ ] Create `circularBuffer` struct in `logMonitor.go`
+- [ ] Implement `Write()` method for circular buffer
+- [ ] Implement `GetHistory()` method for circular buffer
+- [ ] Update `LogMonitor` struct to use new buffer
+- [ ] Update `NewLogMonitorWriter()` to initialize new buffer
+- [ ] Update `LogMonitor.Write()` to use new buffer
+- [ ] Update `LogMonitor.GetHistory()` to use new buffer
+- [ ] Remove `"container/ring"` import
+- [ ] Run `make test-dev` to verify existing tests pass
+- [ ] Add wrap-around test case
+- [ ] Run `make test-all` for final validation

cmd/wol-proxy/index.html

@@ -0,0 +1,64 @@
+<!DOCTYPE html>
+<html>
+<head>
+<meta charset="utf-8">
+<title>Loading...</title>
+<style>
+body {
+    font-family: sans-serif;
+    display: flex;
+    justify-content: center;
+    align-items: center;
+    height: 100vh;
+    margin: 0;
+    background: #f5f5f5;
+}
+.loader {
+    text-align: center;
+}
+.stats {
+    font-size: 18px;
+    color: #333;
+    margin: 20px 0;
+}
+.stats-label {
+    color: #666;
+    font-size: 14px;
+}
+</style>
+</head>
+<body>
+<div class="loader">
+    <p>Waking up upstream server...</p>
+    <div class="stats">
+        <div><span class="stats-label">Time elapsed:</span> <span id="elapsed">0s</span></div>
+        <div><span id="attempts">&nbsp;</span></div>
+    </div>
+</div>
+<script>
+var startTime = Date.now();
+var attempts = 0;
+
+setInterval(function() {
+    var elapsed = (Date.now() - startTime) / 1000;
+    document.getElementById('elapsed').textContent = elapsed.toFixed(1) + 's';
+}, 100);
+
+// Check status every second
+setInterval(function() {
+    attempts++;
+    var dots = '.'.repeat((attempts % 10) || 10);
+    document.getElementById('attempts').textContent = dots;
+
+    fetch('/status')
+        .then(function(r) { return r.text(); })
+        .then(function(t) {
+            if (t.indexOf('status: ready') !== -1) {
+                location.reload();
+            }
+        })
+        .catch(function() {});
+}, 1000);
+</script>
+</body>
+</html>

cmd/wol-proxy/wol-proxy.go

@@ -1,7 +1,9 @@
 package main
 
 import (
+	"bufio"
 	"context"
+	_ "embed"
 	"errors"
 	"flag"
 	"fmt"
@@ -13,10 +15,14 @@ import (
 	"net/url"
 	"os"
 	"os/signal"
+	"strings"
 	"sync"
 	"time"
 )
 
+//go:embed index.html
+var loadingPageHTML string
+
 var (
 	flagMac      = flag.String("mac", "", "mac address to send WoL packet to")
 	flagUpstream = flag.String("upstream", "", "upstream proxy address to send requests to")
@@ -92,14 +98,9 @@ func main() {
 	}()
 
 	// graceful shutdown
-	ctx, stop := signal.NotifyContext(context.Background(), os.Interrupt)
-	defer stop()
+	ctx, _ := signal.NotifyContext(context.Background(), os.Interrupt)
 	<-ctx.Done()
-	shutdownCtx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
-	defer cancel()
-	if err := server.Shutdown(shutdownCtx); err != nil {
-		slog.Error("server shutdown error", "error", err)
-	}
+	server.Close()
 }
 
 type upstreamStatus string
@@ -124,37 +125,86 @@ func newProxy(url *url.URL) *proxyServer {
 		failCount:     0,
 	}
 
-	// start a goroutien to check upstream status
+	// start a goroutine to monitor upstream status via SSE
 	go func() {
-		checkUrl := url.Scheme + "://" + url.Host + "/wol-health"
-		client := &http.Client{Timeout: time.Second}
-		ticker := time.NewTicker(2 * time.Second)
-		defer ticker.Stop()
-		for range ticker.C {
+		eventsUrl := url.Scheme + "://" + url.Host + "/api/events"
+		client := &http.Client{
+			Timeout: 0, // No timeout for SSE connection
+		}
 
-			slog.Debug("checking upstream status at", "url", checkUrl)
-			resp, err := client.Get(checkUrl)
+		waitDuration := 10 * time.Second
 
-			// drain the body
-			if err == nil && resp != nil {
+		for {
+			slog.Debug("connecting to SSE endpoint", "url", eventsUrl)
+
+			req, err := http.NewRequest("GET", eventsUrl, nil)
+			if err != nil {
+				slog.Warn("failed to create SSE request", "error", err)
+				proxy.setStatus(notready)
+				proxy.incFail(1)
+				time.Sleep(waitDuration)
+				continue
+			}
+
+			req.Header.Set("Accept", "text/event-stream")
+			req.Header.Set("Cache-Control", "no-cache")
+			req.Header.Set("Connection", "keep-alive")
+
+			resp, err := client.Do(req)
+			if err != nil {
+				slog.Error("failed to connect to SSE endpoint", "error", err)
+				proxy.setStatus(notready)
+				proxy.incFail(1)
+				time.Sleep(10 * time.Second)
+				continue
+			}
+
+			if resp.StatusCode != http.StatusOK {
+				slog.Warn("SSE endpoint returned non-OK status", "status", resp.StatusCode)
 				_, _ = io.Copy(io.Discard, resp.Body)
 				_ = resp.Body.Close()
-			}
-
-			if err == nil && resp != nil && resp.StatusCode == http.StatusOK {
-				slog.Debug("upstream status: ready")
-				proxy.setStatus(ready)
-				proxy.statusMutex.Lock()
-				proxy.failCount = 0
-				proxy.statusMutex.Unlock()
-			} else {
-				slog.Debug("upstream status: notready", "error", err)
 				proxy.setStatus(notready)
-				proxy.statusMutex.Lock()
-				proxy.failCount++
-				proxy.statusMutex.Unlock()
+				proxy.incFail(1)
+				time.Sleep(10 * time.Second)
+				continue
 			}
 
+			// Successfully connected to SSE endpoint
+			slog.Info("connected to SSE endpoint, upstream ready")
+			proxy.setStatus(ready)
+			proxy.resetFailures()
+
+			// Read from the SSE stream to detect disconnection
+			scanner := bufio.NewScanner(resp.Body)
+
+			// use a fairly large buffer to avoid scanner errors when reading large SSE events
+			buf := make([]byte, 0, 1024*1024*2)
+			scanner.Buffer(buf, 1024*1024*2)
+			events := 0
+			if slog.Default().Enabled(context.Background(), slog.LevelDebug) {
+				fmt.Print("Events: ")
+			}
+			for scanner.Scan() {
+				if slog.Default().Enabled(context.Background(), slog.LevelDebug) {
+					// Just read the events to keep connection alive
+					// We don't need to process the event data
+					events++
+					fmt.Printf("%d, ", events)
+				}
+			}
+			fmt.Println()
+			if err := scanner.Err(); err != nil {
+				slog.Error("error reading from SSE stream", "error", err)
+			}
+
+			// Connection closed or error occurred
+			_ = resp.Body.Close()
+			slog.Info("SSE connection closed, upstream not ready")
+			proxy.setStatus(notready)
+			proxy.incFail(1)
+
+			// Wait before reconnecting
+			time.Sleep(waitDuration)
 		}
 	}()
 
@@ -163,10 +213,8 @@ func newProxy(url *url.URL) *proxyServer {
 
 func (p *proxyServer) ServeHTTP(w http.ResponseWriter, r *http.Request) {
 	if r.Method == "GET" && r.URL.Path == "/status" {
-		p.statusMutex.RLock()
-		status := string(p.status)
-		failCount := p.failCount
-		p.statusMutex.RUnlock()
+		status := string(p.getStatus())
+		failCount := p.getFailures()
 		w.Header().Set("Content-Type", "text/plain")
 		w.WriteHeader(200)
 		fmt.Fprintf(w, "status: %s\n", status)
@@ -175,10 +223,27 @@ func (p *proxyServer) ServeHTTP(w http.ResponseWriter, r *http.Request) {
 	}
 
 	if p.getStatus() == notready {
-		slog.Info("upstream not ready, sending magic packet", "mac", *flagMac)
+		path := r.URL.Path
+		if strings.HasPrefix(path, "/api/events") {
+			slog.Debug("Skipping wake up", "req", path)
+			w.WriteHeader(http.StatusNoContent)
+			return
+		}
+
+		slog.Info("upstream not ready, sending magic packet", "req", path, "from", r.RemoteAddr)
 		if err := sendMagicPacket(*flagMac); err != nil {
 			slog.Warn("failed to send magic WoL packet", "error", err)
 		}
+
+		// For root or UI path requests, return loading page with status polling
+		// the web page will do the polling and redirect when ready
+		if path == "/" || strings.HasPrefix(path, "/ui/") {
+			w.Header().Set("Content-Type", "text/html; charset=utf-8")
+			w.WriteHeader(http.StatusOK)
+			fmt.Fprint(w, loadingPageHTML)
+			return
+		}
+
 		ticker := time.NewTicker(250 * time.Millisecond)
 		timeout, cancel := context.WithTimeout(context.Background(), time.Duration(*flagTimeout)*time.Second)
 		defer cancel()
@@ -213,6 +278,24 @@ func (p *proxyServer) setStatus(status upstreamStatus) {
 	p.status = status
 }
 
+func (p *proxyServer) incFail(num int) {
+	p.statusMutex.Lock()
+	defer p.statusMutex.Unlock()
+	p.failCount += num
+}
+
+func (p *proxyServer) getFailures() int {
+	p.statusMutex.RLock()
+	defer p.statusMutex.RUnlock()
+	return p.failCount
+}
+
+func (p *proxyServer) resetFailures() {
+	p.statusMutex.Lock()
+	defer p.statusMutex.Unlock()
+	p.failCount = 0
+}
+
 func sendMagicPacket(macAddr string) error {
 	hwAddr, err := net.ParseMAC(macAddr)
 	if err != nil {

config-schema.json

@@ -0,0 +1,278 @@
+{
+    "$schema": "https://json-schema.org/draft-07/schema#",
+    "$id": "llama-swap-config-schema.json",
+    "title": "llama-swap configuration",
+    "description": "Configuration file for llama-swap",
+    "type": "object",
+    "required": [
+        "models"
+    ],
+    "definitions": {
+        "macros": {
+            "type": "object",
+            "additionalProperties": {
+                "oneOf": [
+                    {
+                        "type": "string",
+                        "minLength": 0,
+                        "maxLength": 1024
+                    },
+                    {
+                        "type": "number"
+                    },
+                    {
+                        "type": "boolean"
+                    }
+                ]
+            },
+            "propertyNames": {
+                "type": "string",
+                "minLength": 1,
+                "maxLength": 64,
+                "pattern": "^[a-zA-Z0-9_-]+$",
+                "not": {
+                    "enum": [
+                        "PORT",
+                        "MODEL_ID"
+                    ]
+                }
+            },
+            "default": {},
+            "description": "A dictionary of string substitutions. Macros are reusable snippets used in model cmd, cmdStop, proxy, checkEndpoint, filters.stripParams. Macro names must be <64 chars, match ^[a-zA-Z0-9_-]+$, and not be PORT or MODEL_ID. Values can be string, number, or boolean. Macros can reference other macros defined before them."
+        }
+    },
+    "properties": {
+        "healthCheckTimeout": {
+            "type": "integer",
+            "minimum": 15,
+            "default": 120,
+            "description": "Number of seconds to wait for a model to be ready to serve requests."
+        },
+        "logLevel": {
+            "type": "string",
+            "enum": [
+                "debug",
+                "info",
+                "warn",
+                "error"
+            ],
+            "default": "info",
+            "description": "Sets the logging value. Valid values: debug, info, warn, error."
+        },
+        "logTimeFormat": {
+            "type": "string",
+            "enum": [
+                "",
+                "ansic",
+                "unixdate",
+                "rubydate",
+                "rfc822",
+                "rfc822z",
+                "rfc850",
+                "rfc1123",
+                "rfc1123z",
+                "rfc3339",
+                "rfc3339nano",
+                "kitchen",
+                "stamp",
+                "stampmilli",
+                "stampmicro",
+                "stampnano"
+            ],
+            "default": "",
+            "description": "Enables and sets the logging timestamp format. Valid values: \"\", \"ansic\", \"unixdate\", \"rubydate\", \"rfc822\", \"rfc822z\", \"rfc850\", \"rfc1123\", \"rfc1123z\", \"rfc3339\", \"rfc3339nano\", \"kitchen\", \"stamp\", \"stampmilli\", \"stampmicro\", and \"stampnano\". For more info, read: https://pkg.go.dev/time#pkg-constants"
+        },
+        "metricsMaxInMemory": {
+            "type": "integer",
+            "default": 1000,
+            "description": "Maximum number of metrics to keep in memory. Controls how many metrics are stored before older ones are discarded."
+        },
+        "startPort": {
+            "type": "integer",
+            "default": 5800,
+            "description": "Starting port number for the automatic ${PORT} macro. The ${PORT} macro is incremented for every model that uses it."
+        },
+        "sendLoadingState": {
+            "type": "boolean",
+            "default": false,
+            "description": "Inject loading status updates into the reasoning field. When true, a stream of loading messages will be sent to the client."
+        },
+        "includeAliasesInList": {
+            "type": "boolean",
+            "default": false,
+            "description": "Present aliases within the /v1/models OpenAI API listing. when true, model aliases will be output to the API model listing duplicating all fields except for Id so chat UIs can use the alias equivalent to the original."
+        },
+        "macros": {
+            "$ref": "#/definitions/macros"
+        },
+        "models": {
+            "type": "object",
+            "description": "A dictionary of model configurations. Each key is a model's ID. Model settings have defaults if not defined. The model's ID is available as ${MODEL_ID}.",
+            "additionalProperties": {
+                "type": "object",
+                "required": [
+                    "cmd"
+                ],
+                "properties": {
+                    "macros": {
+                        "$ref": "#/definitions/macros"
+                    },
+                    "cmd": {
+                        "type": "string",
+                        "minLength": 1,
+                        "description": "Command to run to start the inference server. Macros can be used. Comments allowed with |."
+                    },
+                    "cmdStop": {
+                        "type": "string",
+                        "default": "",
+                        "description": "Command to run to stop the model gracefully. Uses ${PID} macro for upstream process id. If empty, default shutdown behavior is used."
+                    },
+                    "name": {
+                        "type": "string",
+                        "default": "",
+                        "maxLength": 128,
+                        "description": "Display name for the model. Used in v1/models API response."
+                    },
+                    "description": {
+                        "type": "string",
+                        "default": "",
+                        "maxLength": 1024,
+                        "description": "Description for the model. Used in v1/models API response."
+                    },
+                    "env": {
+                        "type": "array",
+                        "items": {
+                            "type": "string",
+                            "pattern": "^[A-Z_][A-Z0-9_]*=.*$"
+                        },
+                        "default": [],
+                        "description": "Array of environment variables to inject into cmd's environment. Each value is a string in ENV_NAME=value format."
+                    },
+                    "proxy": {
+                        "type": "string",
+                        "default": "http://localhost:${PORT}",
+                        "format": "uri",
+                        "description": "URL where llama-swap routes API requests. If custom port is used in cmd, this must be set."
+                    },
+                    "aliases": {
+                        "type": "array",
+                        "items": {
+                            "type": "string",
+                            "minLength": 1
+                        },
+                        "default": [],
+                        "description": "Alternative model names for this configuration. Must be unique globally."
+                    },
+                    "checkEndpoint": {
+                        "type": "string",
+                        "default": "/health",
+                        "pattern": "^/.*$|^none$",
+                        "description": "URL path to check if the server is ready. Use 'none' to skip health checking."
+                    },
+                    "ttl": {
+                        "type": "integer",
+                        "minimum": 0,
+                        "default": 0,
+                        "description": "Automatically unload the model after ttl seconds. 0 disables unloading. Must be >0 to enable."
+                    },
+                    "useModelName": {
+                        "type": "string",
+                        "default": "",
+                        "description": "Override the model name sent to upstream server. Useful if upstream expects a different name."
+                    },
+                    "filters": {
+                        "type": "object",
+                        "properties": {
+                            "stripParams": {
+                                "type": "string",
+                                "default": "",
+                                "pattern": "^[a-zA-Z0-9_, ]*$",
+                                "description": "Comma separated list of parameters to remove from the request. Used for server-side enforcement of sampling parameters."
+                            }
+                        },
+                        "additionalProperties": false,
+                        "default": {},
+                        "description": "Dictionary of filter settings. Only stripParams is supported."
+                    },
+                    "metadata": {
+                        "type": "object",
+                        "additionalProperties": true,
+                        "default": {},
+                        "description": "Dictionary of arbitrary values included in /v1/models. Can contain complex types. Only passed through in /v1/models responses."
+                    },
+                    "concurrencyLimit": {
+                        "type": "integer",
+                        "minimum": 0,
+                        "default": 0,
+                        "description": "Overrides allowed number of active parallel requests to a model. 0 uses internal default of 10. >0 overrides default. Requests exceeding limit get HTTP 429."
+                    },
+                    "sendLoadingState": {
+                        "type": "boolean",
+                        "description": "Overrides the global sendLoadingState for this model. Ommitting this property will use the global setting."
+                    },
+                    "unlisted": {
+                        "type": "boolean",
+                        "default": false,
+                        "description": "If true the model will not show up in /v1/models responses. It can still be used as normal in API requests."
+                    }
+                }
+            }
+        },
+        "groups": {
+            "type": "object",
+            "additionalProperties": {
+                "type": "object",
+                "required": [
+                    "members"
+                ],
+                "properties": {
+                    "swap": {
+                        "type": "boolean",
+                        "default": true,
+                        "description": "Controls model swapping behaviour within the group. True: only one model runs at a time. False: all models can run together."
+                    },
+                    "exclusive": {
+                        "type": "boolean",
+                        "default": true,
+                        "description": "Controls how the group affects other groups. True: causes all other groups to unload when this group runs a model. False: does not affect other groups."
+                    },
+                    "persistent": {
+                        "type": "boolean",
+                        "default": false,
+                        "description": "Prevents other groups from unloading the models in this group. Does not affect individual model behaviour."
+                    },
+                    "members": {
+                        "type": "array",
+                        "items": {
+                            "type": "string"
+                        },
+                        "description": "Array of model IDs that are members of this group. Model IDs must be defined in models."
+                    }
+                }
+            },
+            "description": "A dictionary of group settings. Provides advanced controls over model swapping behaviour. Model IDs must be defined in models. A model can only be a member of one group. Behaviour controlled via swap, exclusive, persistent."
+        },
+        "hooks": {
+            "type": "object",
+            "properties": {
+                "on_startup": {
+                    "type": "object",
+                    "properties": {
+                        "preload": {
+                            "type": "array",
+                            "items": {
+                                "type": "string"
+                            },
+                            "default": [],
+                            "description": "List of model IDs to load on startup. Model names must match keys in models. When preloading multiple models, define a group to prevent swapping."
+                        }
+                    },
+                    "additionalProperties": false,
+                    "description": "Actions to perform on startup. Only supported action is preload."
+                }
+            },
+            "additionalProperties": false,
+            "description": "A dictionary of event triggers and actions. Only supported hook is on_startup."
+        }
+    }
+}

config.example.yaml

@@ -1,3 +1,6 @@
+# add this modeline for validation in vscode
+# yaml-language-server: $schema=https://raw.githubusercontent.com/mostlygeek/llama-swap/refs/heads/main/config-schema.json
+#
 # llama-swap YAML configuration example
 # -------------------------------------
 #
@@ -23,6 +26,24 @@ healthCheckTimeout: 500
 # - Valid log levels: debug, info, warn, error
 logLevel: info
 
+# logTimeFormat: enables and sets the logging timestamp format
+# - optional, default (disabled): ""
+# - Valid values: "", "ansic", "unixdate", "rubydate", "rfc822", "rfc822z",
+#   "rfc850", "rfc1123", "rfc1123z", "rfc3339", "rfc3339nano", "kitchen",
+#   "stamp", "stampmilli", "stampmicro", and "stampnano".
+# - For more info, read: https://pkg.go.dev/time#pkg-constants
+logTimeFormat: ""
+
+# logToStdout: controls what is logged to stdout
+# - optional, default: "proxy"
+# - valid values:
+#   - "proxy": logs generated by llama-swap when swapping models,
+#      handling requests, etc.
+#   - "upstream": a copy of an upstream processes stdout logs
+#   - "both": both the proxy and upstream logs interleaved together
+#   - "none": no logs are ever written to stdout
+logToStdout: "proxy"
+
 # metricsMaxInMemory: maximum number of metrics to keep in memory
 # - optional, default: 1000
 # - controls how many metrics are stored in memory before older ones are discarded
@@ -35,6 +56,20 @@ metricsMaxInMemory: 1000
 # - it is automatically incremented for every model that uses it
 startPort: 10001
 
+# sendLoadingState: inject loading status updates into the reasoning (thinking)
+# field
+# - optional, default: false
+# - when true, a stream of loading messages will be sent to the client in the
+#   reasoning field so chat UIs can show that loading is in progress.
+# - see #366 for more details
+sendLoadingState: true
+
+# includeAliasesInList: present aliases within the /v1/models OpenAI API listing
+# - optional, default: false
+# - when true, model aliases will be output to the API model listing duplicating
+#   all fields except for Id so chat UIs can use the alias equivalent to the original.
+includeAliasesInList: false
+
 # macros: a dictionary of string substitutions
 # - optional, default: empty dictionary
 # - macros are reusable snippets
@@ -64,7 +99,6 @@ macros:
 # - the model's ID is available in the ${MODEL_ID} macro, also available in macros defined above
 # - below are examples of the all the settings a model can have
 models:
-
   # keys are the model names used in API requests
   "llama":
     # macros: a dictionary of string substitutions specific to this model
@@ -184,6 +218,10 @@ models:
     # - recommended to be omitted and the default used
     concurrencyLimit: 0
 
+    # sendLoadingState: overrides the global sendLoadingState setting for this model
+    # - optional, default: undefined (use global setting)
+    sendLoadingState: false
+
   # Unlisted model example:
   "qwen-unlisted":
     # unlisted: boolean, true or false
@@ -286,10 +324,10 @@ hooks:
   # - optional, default: empty dictionary
   # - the only supported action is preload
   on_startup:
-        # preload: a list of model ids to load on startup
-        # - optional, default: empty list
-        # - model names must match keys in the models sections
-        # - when preloading multiple models at once, define a group
-        #   otherwise models will be loaded and swapped out
+    # preload: a list of model ids to load on startup
+    # - optional, default: empty list
+    # - model names must match keys in the models sections
+    # - when preloading multiple models at once, define a group
+    #   otherwise models will be loaded and swapped out
     preload:
-      - "llama"
+      - "llama"

docker/build-container.sh

@@ -20,36 +20,60 @@ if [[ -z "$GITHUB_TOKEN" ]]; then
   exit 1
 fi
 
+# Set llama.cpp base image, customizable using the BASE_LLAMACPP_IMAGE environment
+# variable, this permits testing with forked llama.cpp repositories
+BASE_IMAGE=${BASE_LLAMACPP_IMAGE:-ghcr.io/ggml-org/llama.cpp}
+
+# Set llama-swap repository, automatically uses GITHUB_REPOSITORY variable
+# to enable easy container builds on forked repos
+LS_REPO=${GITHUB_REPOSITORY:-mostlygeek/llama-swap}
+
 # the most recent llama-swap tag
 # have to strip out the 'v' due to .tar.gz file naming
-LS_VER=$(curl -s https://api.github.com/repos/mostlygeek/llama-swap/releases/latest | jq -r .tag_name | sed 's/v//')
+LS_VER=$(curl -s https://api.github.com/repos/${LS_REPO}/releases/latest | jq -r .tag_name | sed 's/v//')
 
 if [ "$ARCH" == "cpu" ]; then
-    # cpu only containers just use the latest available
-    CONTAINER_LATEST="ghcr.io/mostlygeek/llama-swap:cpu"
-    echo "Building ${CONTAINER_LATEST} $LS_VER"
-    docker build -f llama-swap.Containerfile --build-arg BASE_TAG=server --build-arg LS_VER=${LS_VER} -t ${CONTAINER_LATEST} .
-    if [ "$PUSH_IMAGES" == "true" ]; then
-      docker push ${CONTAINER_LATEST}
-    fi
+    # cpu only containers just use the server tag
+    LCPP_TAG=$(curl -s -H "Authorization: Bearer $GITHUB_TOKEN" \
+        "https://api.github.com/users/ggml-org/packages/container/llama.cpp/versions" \
+        | jq -r '.[] | select(.metadata.container.tags[] | startswith("server")) | .metadata.container.tags[]' \
+        | sort -r | head -n1 | awk -F '-' '{print $3}')
+    BASE_TAG=server-${LCPP_TAG}
 else
     LCPP_TAG=$(curl -s -H "Authorization: Bearer $GITHUB_TOKEN" \
         "https://api.github.com/users/ggml-org/packages/container/llama.cpp/versions" \
         | jq -r --arg arch "$ARCH" '.[] | select(.metadata.container.tags[] | startswith("server-\($arch)")) | .metadata.container.tags[]' \
         | sort -r | head -n1 | awk -F '-' '{print $3}')
+    BASE_TAG=server-${ARCH}-${LCPP_TAG}
+fi
 
-    # Abort if LCPP_TAG is empty.
-    if [[ -z "$LCPP_TAG" ]]; then
-        echo "Abort: Could not find llama-server container for arch: $ARCH"
-        exit 1
-    fi
+# Abort if LCPP_TAG is empty.
+if [[ -z "$LCPP_TAG" ]]; then
+    echo "Abort: Could not find llama-server container for arch: $ARCH"
+    exit 1
+fi
 
-    CONTAINER_TAG="ghcr.io/mostlygeek/llama-swap:v${LS_VER}-${ARCH}-${LCPP_TAG}"
-    CONTAINER_LATEST="ghcr.io/mostlygeek/llama-swap:${ARCH}"
-    echo "Building ${CONTAINER_TAG} $LS_VER"
-    docker build -f llama-swap.Containerfile --build-arg BASE_TAG=server-${ARCH}-${LCPP_TAG} --build-arg LS_VER=${LS_VER} -t ${CONTAINER_TAG} -t ${CONTAINER_LATEST} .
-    if [ "$PUSH_IMAGES" == "true" ]; then
-      docker push ${CONTAINER_TAG}
-      docker push ${CONTAINER_LATEST}
-    fi
-fi
+for CONTAINER_TYPE in non-root root; do
+  CONTAINER_TAG="ghcr.io/${LS_REPO}:v${LS_VER}-${ARCH}-${LCPP_TAG}"
+  CONTAINER_LATEST="ghcr.io/${LS_REPO}:${ARCH}"
+  USER_UID=0
+  USER_GID=0
+  USER_HOME=/root
+
+  if [ "$CONTAINER_TYPE" == "non-root" ]; then
+    CONTAINER_TAG="${CONTAINER_TAG}-non-root"
+    CONTAINER_LATEST="${CONTAINER_LATEST}-non-root"
+    USER_UID=10001
+    USER_GID=10001
+    USER_HOME=/app
+  fi
+
+  echo "Building $CONTAINER_TYPE $CONTAINER_TAG $LS_VER"
+  docker build -f llama-swap.Containerfile --build-arg BASE_TAG=${BASE_TAG} --build-arg LS_VER=${LS_VER} --build-arg UID=${USER_UID} \
+    --build-arg LS_REPO=${LS_REPO} --build-arg GID=${USER_GID} --build-arg USER_HOME=${USER_HOME} -t ${CONTAINER_TAG} -t ${CONTAINER_LATEST} \
+    --build-arg BASE_IMAGE=${BASE_IMAGE} .
+  if [ "$PUSH_IMAGES" == "true" ]; then
+    docker push ${CONTAINER_TAG}
+    docker push ${CONTAINER_LATEST}
+  fi
+done

docker/llama-swap.Containerfile

@@ -1,16 +1,44 @@
+ARG BASE_IMAGE=ghcr.io/ggml-org/llama.cpp
 ARG BASE_TAG=server-cuda
-FROM ghcr.io/ggml-org/llama.cpp:${BASE_TAG}
+FROM ${BASE_IMAGE}:${BASE_TAG}
 
 # has to be after the FROM
-ARG LS_VER=89
+ARG LS_VER=170
+ARG LS_REPO=mostlygeek/llama-swap
+
+# Set default UID/GID arguments
+ARG UID=10001
+ARG GID=10001
+ARG USER_HOME=/app
+
+# Add user/group
+ENV HOME=$USER_HOME
+RUN if [ $UID -ne 0 ]; then \
+      if [ $GID -ne 0 ]; then \
+        groupadd --system --gid $GID app; \
+      fi; \
+      useradd --system --uid $UID --gid $GID \
+      --home $USER_HOME app; \
+    fi
+
+# Handle paths
+RUN mkdir --parents $HOME /app
+RUN chown --recursive $UID:$GID $HOME /app
+
+# Switch user
+USER $UID:$GID
 
 WORKDIR /app
-RUN \
-    curl -LO https://github.com/mostlygeek/llama-swap/releases/download/v"${LS_VER}"/llama-swap_"${LS_VER}"_linux_amd64.tar.gz && \
-    tar -zxf llama-swap_"${LS_VER}"_linux_amd64.tar.gz && \
-    rm llama-swap_"${LS_VER}"_linux_amd64.tar.gz
 
-COPY config.example.yaml /app/config.yaml
+# Add /app to PATH
+ENV PATH="/app:${PATH}"
+
+RUN \
+    curl -LO "https://github.com/${LS_REPO}/releases/download/v${LS_VER}/llama-swap_${LS_VER}_linux_amd64.tar.gz" && \
+    tar -zxf "llama-swap_${LS_VER}_linux_amd64.tar.gz" && \
+    rm "llama-swap_${LS_VER}_linux_amd64.tar.gz"
+
+COPY --chown=$UID:$GID config.example.yaml /app/config.yaml
 
 HEALTHCHECK CMD curl -f http://localhost:8080/ || exit 1
-ENTRYPOINT [ "/app/llama-swap", "-config", "/app/config.yaml" ]
+ENTRYPOINT [ "/app/llama-swap", "-config", "/app/config.yaml" ]

docs/configuration.md

@@ -89,6 +89,9 @@ llama-swap supports many more features to customize how you want to manage your
 > This is a copy of `config.example.yaml`. Always check that for the most up to date examples.
 
 ```yaml
+# add this modeline for validation in vscode
+# yaml-language-server: $schema=https://raw.githubusercontent.com/mostlygeek/llama-swap/refs/heads/main/config-schema.json
+#
 # llama-swap YAML configuration example
 # -------------------------------------
 #
@@ -114,6 +117,24 @@ healthCheckTimeout: 500
 # - Valid log levels: debug, info, warn, error
 logLevel: info
 
+# logTimeFormat: enables and sets the logging timestamp format
+# - optional, default (disabled): ""
+# - Valid values: "", "ansic", "unixdate", "rubydate", "rfc822", "rfc822z",
+#   "rfc850", "rfc1123", "rfc1123z", "rfc3339", "rfc3339nano", "kitchen",
+#   "stamp", "stampmilli", "stampmicro", and "stampnano".
+# - For more info, read: https://pkg.go.dev/time#pkg-constants
+logTimeFormat: ""
+
+# logToStdout: controls what is logged to stdout
+# - optional, default: "proxy"
+# - valid values:
+#   - "proxy": logs generated by llama-swap when swapping models,
+#      handling requests, etc.
+#   - "upstream": a copy of an upstream processes stdout logs
+#   - "both": both the proxy and upstream logs interleaved together
+#   - "none": no logs are ever written to stdout
+logToStdout: "proxy"
+
 # metricsMaxInMemory: maximum number of metrics to keep in memory
 # - optional, default: 1000
 # - controls how many metrics are stored in memory before older ones are discarded
@@ -126,6 +147,20 @@ metricsMaxInMemory: 1000
 # - it is automatically incremented for every model that uses it
 startPort: 10001
 
+# sendLoadingState: inject loading status updates into the reasoning (thinking)
+# field
+# - optional, default: false
+# - when true, a stream of loading messages will be sent to the client in the
+#   reasoning field so chat UIs can show that loading is in progress.
+# - see #366 for more details
+sendLoadingState: true
+
+# includeAliasesInList: present aliases within the /v1/models OpenAI API listing
+# - optional, default: false
+# - when true, model aliases will be output to the API model listing duplicating
+#   all fields except for Id so chat UIs can use the alias equivalent to the original.
+includeAliasesInList: false
+
 # macros: a dictionary of string substitutions
 # - optional, default: empty dictionary
 # - macros are reusable snippets
@@ -274,6 +309,10 @@ models:
     # - recommended to be omitted and the default used
     concurrencyLimit: 0
 
+    # sendLoadingState: overrides the global sendLoadingState setting for this model
+    # - optional, default: undefined (use global setting)
+    sendLoadingState: false
+
   # Unlisted model example:
   "qwen-unlisted":
     # unlisted: boolean, true or false

docs/container-security.md

@@ -0,0 +1,9 @@
+## Container Security
+
+For convenience, the default container images use the **root** user within the container. This permits simplified access to host resources including volume mounts and hardware devices under `/dev/dri` (_for Vulkan support_). But this can widen the attack surface to privilege escalation exploits.
+
+Alternative images, tagged as `non-root`, are also available. For example, `llama-swap:cpu-non-root` uses the unprivileged **app** user by default. Depending on deployment requirements, additional configuration may be necessary to ensure that the container retains access to required hosts resources. This might entail customizing host filesystem permissions/ownership appropriately or injecting host group membership into the container.
+
+Docker offers a [system-wide option enabling user namespace remapping](https://docs.docker.com/engine/security/userns-remap/) to accommodate situations were a **root** container user is required but also mentions that _"The best way to prevent privilege-escalation attacks from within a container is to configure your container's applications to run as unprivileged users."_ Podman offers similar capability, per-container, to [set UID/GID mapping in a new user namespace](https://docs.podman.io/en/latest/markdown/podman-run.1.html#set-uid-gid-mapping-in-a-new-user-namespace).
+
+The Large Language Model (_LLM/AI_) ecosystem is rapidly evolving and [serious security vulnerabilities have surfaced in the past](https://huggingface.co/docs/hub/security-pickle). These alternative _non-root_ images could reduce the impact of future unknown problems. However, proper planning and configuration is recommended to utilize them.

go.mod

@@ -1,6 +1,6 @@
 module github.com/mostlygeek/llama-swap
 
-go 1.23.0
+go 1.25.4
 
 require (
 	github.com/billziss-gh/golib v0.2.0
@@ -37,9 +37,9 @@ require (
 	github.com/twitchyliquid64/golang-asm v0.15.1 // indirect
 	github.com/ugorji/go/codec v1.2.12 // indirect
 	golang.org/x/arch v0.8.0 // indirect
-	golang.org/x/crypto v0.36.0 // indirect
-	golang.org/x/net v0.38.0 // indirect
-	golang.org/x/sys v0.31.0 // indirect
-	golang.org/x/text v0.23.0 // indirect
+	golang.org/x/crypto v0.45.0 // indirect
+	golang.org/x/net v0.47.0 // indirect
+	golang.org/x/sys v0.38.0 // indirect
+	golang.org/x/text v0.31.0 // indirect
 	google.golang.org/protobuf v1.34.1 // indirect
 )

go.sum

@@ -80,16 +80,16 @@ github.com/ugorji/go/codec v1.2.12/go.mod h1:UNopzCgEMSXjBc6AOMqYvWC1ktqTAfzJZUZ
 golang.org/x/arch v0.0.0-20210923205945-b76863e36670/go.mod h1:5om86z9Hs0C8fWVUuoMHwpExlXzs5Tkyp9hOrfG7pp8=
 golang.org/x/arch v0.8.0 h1:3wRIsP3pM4yUptoR96otTUOXI367OS0+c9eeRi9doIc=
 golang.org/x/arch v0.8.0/go.mod h1:FEVrYAQjsQXMVJ1nsMoVVXPZg6p2JE2mx8psSWTDQys=
-golang.org/x/crypto v0.36.0 h1:AnAEvhDddvBdpY+uR+MyHmuZzzNqXSe/GvuDeob5L34=
-golang.org/x/crypto v0.36.0/go.mod h1:Y4J0ReaxCR1IMaabaSMugxJES1EpwhBHhv2bDHklZvc=
-golang.org/x/net v0.38.0 h1:vRMAPTMaeGqVhG5QyLJHqNDwecKTomGeqbnfZyKlBI8=
-golang.org/x/net v0.38.0/go.mod h1:ivrbrMbzFq5J41QOQh0siUuly180yBYtLp+CKbEaFx8=
+golang.org/x/crypto v0.45.0 h1:jMBrvKuj23MTlT0bQEOBcAE0mjg8mK9RXFhRH6nyF3Q=
+golang.org/x/crypto v0.45.0/go.mod h1:XTGrrkGJve7CYK7J8PEww4aY7gM3qMCElcJQ8n8JdX4=
+golang.org/x/net v0.47.0 h1:Mx+4dIFzqraBXUugkia1OOvlD6LemFo1ALMHjrXDOhY=
+golang.org/x/net v0.47.0/go.mod h1:/jNxtkgq5yWUGYkaZGqo27cfGZ1c5Nen03aYrrKpVRU=
 golang.org/x/sys v0.5.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.6.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
-golang.org/x/sys v0.31.0 h1:ioabZlmFYtWhL+TRYpcnNlLwhyxaM9kWTDEmfnprqik=
-golang.org/x/sys v0.31.0/go.mod h1:BJP2sWEmIv4KK5OTEluFJCKSidICx8ciO85XgH3Ak8k=
-golang.org/x/text v0.23.0 h1:D71I7dUrlY+VX0gQShAThNGHFxZ13dGLBHQLVl1mJlY=
-golang.org/x/text v0.23.0/go.mod h1:/BLNzu4aZCJ1+kcD0DNRotWKage4q2rGVAg4o22unh4=
+golang.org/x/sys v0.38.0 h1:3yZWxaJjBmCWXqhN1qh02AkOnCQ1poK6oF+a7xWL6Gc=
+golang.org/x/sys v0.38.0/go.mod h1:OgkHotnGiDImocRcuBABYBEXf8A9a87e/uXjp9XT3ks=
+golang.org/x/text v0.31.0 h1:aC8ghyu4JhP8VojJ2lEHBnochRno1sgL6nEi9WGFGMM=
+golang.org/x/text v0.31.0/go.mod h1:tKRAlv61yKIjGGHX/4tP1LTbc13YSec1pxVEWXzfoeM=
 golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543 h1:E7g+9GITq07hpfrRu66IVDexMakfv52eLZ2CXBWiKr4=
 golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
 google.golang.org/protobuf v1.34.1 h1:9ddQBjfCyZPOHPUiPxpYESBLc+T8P3E+Vo4IbKZgFWg=

llama-swap.go

@@ -95,7 +95,9 @@ func main() {
 
 			fmt.Println("Configuration Changed")
 			currentPM.Shutdown()
-			srv.Handler = proxy.New(conf)
+			newPM := proxy.New(conf)
+			newPM.SetVersion(date, commit, version)
+			srv.Handler = newPM
 			fmt.Println("Configuration Reloaded")
 
 			// wait a few seconds and tell any UI to reload
@@ -110,7 +112,9 @@ func main() {
 				fmt.Printf("Error, unable to load configuration: %v\n", err)
 				os.Exit(1)
 			}
-			srv.Handler = proxy.New(conf)
+			newPM := proxy.New(conf)
+			newPM.SetVersion(date, commit, version)
+			srv.Handler = newPM
 		}
 	}
 

proxy/config/config.go

@@ -15,6 +15,12 @@ import (
 )
 
 const DEFAULT_GROUP_ID = "(default)"
+const (
+	LogToStdoutProxy    = "proxy"
+	LogToStdoutUpstream = "upstream"
+	LogToStdoutBoth     = "both"
+	LogToStdoutNone     = "none"
+)
 
 type MacroEntry struct {
 	Name  string
@@ -113,6 +119,8 @@ type Config struct {
 	HealthCheckTimeout int                    `yaml:"healthCheckTimeout"`
 	LogRequests        bool                   `yaml:"logRequests"`
 	LogLevel           string                 `yaml:"logLevel"`
+	LogTimeFormat      string                 `yaml:"logTimeFormat"`
+	LogToStdout        string                 `yaml:"logToStdout"`
 	MetricsMaxInMemory int                    `yaml:"metricsMaxInMemory"`
 	Models             map[string]ModelConfig `yaml:"models"` /* key is model ID */
 	Profiles           map[string][]string    `yaml:"profiles"`
@@ -129,6 +137,12 @@ type Config struct {
 
 	// hooks, see: #209
 	Hooks HooksConfig `yaml:"hooks"`
+
+	// send loading state in reasoning
+	SendLoadingState bool `yaml:"sendLoadingState"`
+
+	// present aliases to /v1/models OpenAI API listing
+	IncludeAliasesInList bool `yaml:"includeAliasesInList"`
 }
 
 func (c *Config) RealModelName(search string) (string, bool) {
@@ -169,6 +183,8 @@ func LoadConfigFromReader(r io.Reader) (Config, error) {
 		HealthCheckTimeout: 120,
 		StartPort:          5800,
 		LogLevel:           "info",
+		LogTimeFormat:      "",
+		LogToStdout:        LogToStdoutProxy,
 		MetricsMaxInMemory: 1000,
 	}
 	err = yaml.Unmarshal(data, &config)
@@ -185,6 +201,12 @@ func LoadConfigFromReader(r io.Reader) (Config, error) {
 		return Config{}, fmt.Errorf("startPort must be greater than 1")
 	}
 
+	switch config.LogToStdout {
+	case LogToStdoutProxy, LogToStdoutUpstream, LogToStdoutBoth, LogToStdoutNone:
+	default:
+		return Config{}, fmt.Errorf("logToStdout must be one of: proxy, upstream, both, none")
+	}
+
 	// Populate the aliases map
 	config.aliases = make(map[string]string)
 	for modelName, modelConfig := range config.Models {
@@ -350,6 +372,13 @@ func LoadConfigFromReader(r io.Reader) (Config, error) {
 			)
 		}
 
+		// if sendLoadingState is nil, set it to the global config value
+		// see #366
+		if modelConfig.SendLoadingState == nil {
+			v := config.SendLoadingState // copy it
+			modelConfig.SendLoadingState = &v
+		}
+
 		config.Models[modelId] = modelConfig
 	}
 

proxy/config/config_posix_test.go

@@ -58,6 +58,7 @@ models:
 	assert.Equal(t, 120, config.HealthCheckTimeout)
 	assert.Equal(t, 5800, config.StartPort)
 	assert.Equal(t, "info", config.LogLevel)
+	assert.Equal(t, "", config.LogTimeFormat)
 
 	// Test default group exists
 	defaultGroup, exists := config.Groups["(default)"]
@@ -160,9 +161,13 @@ groups:
 		t.Fatalf("Failed to load config: %v", err)
 	}
 
+	modelLoadingState := false
+
 	expected := Config{
-		LogLevel:  "info",
-		StartPort: 5800,
+		LogLevel:      "info",
+		LogTimeFormat: "",
+		LogToStdout:   LogToStdoutProxy,
+		StartPort:     5800,
 		Macros: MacroList{
 			{"svr-path", "path/to/server"},
 		},
@@ -171,36 +176,41 @@ groups:
 				Preload: []string{"model1", "model2"},
 			},
 		},
+		SendLoadingState: false,
 		Models: map[string]ModelConfig{
 			"model1": {
-				Cmd:           "path/to/cmd --arg1 one",
-				Proxy:         "http://localhost:8080",
-				Aliases:       []string{"m1", "model-one"},
-				Env:           []string{"VAR1=value1", "VAR2=value2"},
-				CheckEndpoint: "/health",
-				Name:          "Model 1",
-				Description:   "This is model 1",
+				Cmd:              "path/to/cmd --arg1 one",
+				Proxy:            "http://localhost:8080",
+				Aliases:          []string{"m1", "model-one"},
+				Env:              []string{"VAR1=value1", "VAR2=value2"},
+				CheckEndpoint:    "/health",
+				Name:             "Model 1",
+				Description:      "This is model 1",
+				SendLoadingState: &modelLoadingState,
 			},
 			"model2": {
-				Cmd:           "path/to/server --arg1 one",
-				Proxy:         "http://localhost:8081",
-				Aliases:       []string{"m2"},
-				Env:           []string{},
-				CheckEndpoint: "/",
+				Cmd:              "path/to/server --arg1 one",
+				Proxy:            "http://localhost:8081",
+				Aliases:          []string{"m2"},
+				Env:              []string{},
+				CheckEndpoint:    "/",
+				SendLoadingState: &modelLoadingState,
 			},
 			"model3": {
-				Cmd:           "path/to/cmd --arg1 one",
-				Proxy:         "http://localhost:8081",
-				Aliases:       []string{"mthree"},
-				Env:           []string{},
-				CheckEndpoint: "/",
+				Cmd:              "path/to/cmd --arg1 one",
+				Proxy:            "http://localhost:8081",
+				Aliases:          []string{"mthree"},
+				Env:              []string{},
+				CheckEndpoint:    "/",
+				SendLoadingState: &modelLoadingState,
 			},
 			"model4": {
-				Cmd:           "path/to/cmd --arg1 one",
-				Proxy:         "http://localhost:8082",
-				CheckEndpoint: "/",
-				Aliases:       []string{},
-				Env:           []string{},
+				Cmd:              "path/to/cmd --arg1 one",
+				Proxy:            "http://localhost:8082",
+				CheckEndpoint:    "/",
+				Aliases:          []string{},
+				Env:              []string{},
+				SendLoadingState: &modelLoadingState,
 			},
 		},
 		HealthCheckTimeout: 15,

proxy/config/config_windows_test.go

@@ -55,6 +55,7 @@ models:
 	assert.Equal(t, 120, config.HealthCheckTimeout)
 	assert.Equal(t, 5800, config.StartPort)
 	assert.Equal(t, "info", config.LogLevel)
+	assert.Equal(t, "", config.LogTimeFormat)
 
 	// Test default group exists
 	defaultGroup, exists := config.Groups["(default)"]
@@ -152,44 +153,53 @@ groups:
 		t.Fatalf("Failed to load config: %v", err)
 	}
 
+	modelLoadingState := false
+
 	expected := Config{
-		LogLevel:  "info",
-		StartPort: 5800,
+		LogLevel:      "info",
+		LogTimeFormat: "",
+		LogToStdout:   LogToStdoutProxy,
+		StartPort:     5800,
 		Macros: MacroList{
 			{"svr-path", "path/to/server"},
 		},
+		SendLoadingState: false,
 		Models: map[string]ModelConfig{
 			"model1": {
-				Cmd:           "path/to/cmd --arg1 one",
-				CmdStop:       "taskkill /f /t /pid ${PID}",
-				Proxy:         "http://localhost:8080",
-				Aliases:       []string{"m1", "model-one"},
-				Env:           []string{"VAR1=value1", "VAR2=value2"},
-				CheckEndpoint: "/health",
+				Cmd:              "path/to/cmd --arg1 one",
+				CmdStop:          "taskkill /f /t /pid ${PID}",
+				Proxy:            "http://localhost:8080",
+				Aliases:          []string{"m1", "model-one"},
+				Env:              []string{"VAR1=value1", "VAR2=value2"},
+				CheckEndpoint:    "/health",
+				SendLoadingState: &modelLoadingState,
 			},
 			"model2": {
-				Cmd:           "path/to/server --arg1 one",
-				CmdStop:       "taskkill /f /t /pid ${PID}",
-				Proxy:         "http://localhost:8081",
-				Aliases:       []string{"m2"},
-				Env:           []string{},
-				CheckEndpoint: "/",
+				Cmd:              "path/to/server --arg1 one",
+				CmdStop:          "taskkill /f /t /pid ${PID}",
+				Proxy:            "http://localhost:8081",
+				Aliases:          []string{"m2"},
+				Env:              []string{},
+				CheckEndpoint:    "/",
+				SendLoadingState: &modelLoadingState,
 			},
 			"model3": {
-				Cmd:           "path/to/cmd --arg1 one",
-				CmdStop:       "taskkill /f /t /pid ${PID}",
-				Proxy:         "http://localhost:8081",
-				Aliases:       []string{"mthree"},
-				Env:           []string{},
-				CheckEndpoint: "/",
+				Cmd:              "path/to/cmd --arg1 one",
+				CmdStop:          "taskkill /f /t /pid ${PID}",
+				Proxy:            "http://localhost:8081",
+				Aliases:          []string{"mthree"},
+				Env:              []string{},
+				CheckEndpoint:    "/",
+				SendLoadingState: &modelLoadingState,
 			},
 			"model4": {
-				Cmd:           "path/to/cmd --arg1 one",
-				CmdStop:       "taskkill /f /t /pid ${PID}",
-				Proxy:         "http://localhost:8082",
-				CheckEndpoint: "/",
-				Aliases:       []string{},
-				Env:           []string{},
+				Cmd:              "path/to/cmd --arg1 one",
+				CmdStop:          "taskkill /f /t /pid ${PID}",
+				Proxy:            "http://localhost:8082",
+				CheckEndpoint:    "/",
+				Aliases:          []string{},
+				Env:              []string{},
+				SendLoadingState: &modelLoadingState,
 			},
 		},
 		HealthCheckTimeout: 15,

proxy/config/model_config.go

@@ -35,6 +35,9 @@ type ModelConfig struct {
 	// Metadata: see #264
 	// Arbitrary metadata that can be exposed through the API
 	Metadata map[string]any `yaml:"metadata"`
+
+	// override global setting
+	SendLoadingState *bool `yaml:"sendLoadingState"`
 }
 
 func (m *ModelConfig) UnmarshalYAML(unmarshal func(interface{}) error) error {

proxy/config/model_config_test.go

@@ -50,5 +50,25 @@ models:
 			}
 		})
 	}
-
+}
+
+func TestConfig_ModelSendLoadingState(t *testing.T) {
+	content := `
+sendLoadingState: true
+models:
+  model1:
+    cmd: path/to/cmd --port ${PORT}
+    sendLoadingState: false
+  model2:
+    cmd: path/to/cmd --port ${PORT}
+`
+	config, err := LoadConfigFromReader(strings.NewReader(content))
+	assert.NoError(t, err)
+	assert.True(t, config.SendLoadingState)
+	if assert.NotNil(t, config.Models["model1"].SendLoadingState) {
+		assert.False(t, *config.Models["model1"].SendLoadingState)
+	}
+	if assert.NotNil(t, config.Models["model2"].SendLoadingState) {
+		assert.True(t, *config.Models["model2"].SendLoadingState)
+	}
 }

proxy/logMonitor.go

@@ -1,16 +1,95 @@
 package proxy
 
 import (
-	"container/ring"
 	"context"
 	"fmt"
 	"io"
 	"os"
 	"sync"
+	"time"
 
 	"github.com/mostlygeek/llama-swap/event"
 )
 
+// circularBuffer is a fixed-size circular byte buffer that overwrites
+// oldest data when full. It provides O(1) writes and O(n) reads.
+type circularBuffer struct {
+	data []byte // pre-allocated capacity
+	head int    // next write position
+	size int    // current number of bytes stored (0 to cap)
+}
+
+func newCircularBuffer(capacity int) *circularBuffer {
+	return &circularBuffer{
+		data: make([]byte, capacity),
+		head: 0,
+		size: 0,
+	}
+}
+
+// Write appends bytes to the buffer, overwriting oldest data when full.
+// Data is copied into the internal buffer (not stored by reference).
+func (cb *circularBuffer) Write(p []byte) {
+	if len(p) == 0 {
+		return
+	}
+
+	cap := len(cb.data)
+
+	// If input is larger than capacity, only keep the last cap bytes
+	if len(p) >= cap {
+		copy(cb.data, p[len(p)-cap:])
+		cb.head = 0
+		cb.size = cap
+		return
+	}
+
+	// Calculate how much space is available from head to end of buffer
+	firstPart := cap - cb.head
+	if firstPart >= len(p) {
+		// All data fits without wrapping
+		copy(cb.data[cb.head:], p)
+		cb.head = (cb.head + len(p)) % cap
+	} else {
+		// Data wraps around
+		copy(cb.data[cb.head:], p[:firstPart])
+		copy(cb.data[:len(p)-firstPart], p[firstPart:])
+		cb.head = len(p) - firstPart
+	}
+
+	// Update size
+	cb.size += len(p)
+	if cb.size > cap {
+		cb.size = cap
+	}
+}
+
+// GetHistory returns all buffered data in correct order (oldest to newest).
+// Returns a new slice (copy), not a view into internal buffer.
+func (cb *circularBuffer) GetHistory() []byte {
+	if cb.size == 0 {
+		return nil
+	}
+
+	result := make([]byte, cb.size)
+	cap := len(cb.data)
+
+	// Calculate start position (oldest data)
+	start := (cb.head - cb.size + cap) % cap
+
+	if start+cb.size <= cap {
+		// Data is contiguous, single copy
+		copy(result, cb.data[start:start+cb.size])
+	} else {
+		// Data wraps around, two copies
+		firstPart := cap - start
+		copy(result[:firstPart], cb.data[start:])
+		copy(result[firstPart:], cb.data[:cb.size-firstPart])
+	}
+
+	return result
+}
+
 type LogLevel int
 
 const (
@@ -18,12 +97,14 @@ const (
 	LevelInfo
 	LevelWarn
 	LevelError
+
+	LogBufferSize = 100 * 1024
 )
 
 type LogMonitor struct {
 	eventbus *event.Dispatcher
 	mu       sync.RWMutex
-	buffer   *ring.Ring
+	buffer   *circularBuffer
 	bufferMu sync.RWMutex
 
 	// typically this can be os.Stdout
@@ -32,6 +113,9 @@ type LogMonitor struct {
 	// logging levels
 	level  LogLevel
 	prefix string
+
+	// timestamps
+	timeFormat string
 }
 
 func NewLogMonitor() *LogMonitor {
@@ -40,11 +124,12 @@ func NewLogMonitor() *LogMonitor {
 
 func NewLogMonitorWriter(stdout io.Writer) *LogMonitor {
 	return &LogMonitor{
-		eventbus: event.NewDispatcherConfig(1000),
-		buffer:   ring.New(10 * 1024), // keep 10KB of buffered logs
-		stdout:   stdout,
-		level:    LevelInfo,
-		prefix:   "",
+		eventbus:   event.NewDispatcherConfig(1000),
+		buffer:     nil, // lazy initialized on first Write
+		stdout:     stdout,
+		level:      LevelInfo,
+		prefix:     "",
+		timeFormat: "",
 	}
 }
 
@@ -59,12 +144,15 @@ func (w *LogMonitor) Write(p []byte) (n int, err error) {
 	}
 
 	w.bufferMu.Lock()
-	bufferCopy := make([]byte, len(p))
-	copy(bufferCopy, p)
-	w.buffer.Value = bufferCopy
-	w.buffer = w.buffer.Next()
+	if w.buffer == nil {
+		w.buffer = newCircularBuffer(LogBufferSize)
+	}
+	w.buffer.Write(p)
 	w.bufferMu.Unlock()
 
+	// Make a copy for broadcast to preserve immutability
+	bufferCopy := make([]byte, len(p))
+	copy(bufferCopy, p)
 	w.broadcast(bufferCopy)
 	return n, nil
 }
@@ -72,16 +160,18 @@ func (w *LogMonitor) Write(p []byte) (n int, err error) {
 func (w *LogMonitor) GetHistory() []byte {
 	w.bufferMu.RLock()
 	defer w.bufferMu.RUnlock()
+	if w.buffer == nil {
+		return nil
+	}
+	return w.buffer.GetHistory()
+}
 
-	var history []byte
-	w.buffer.Do(func(p any) {
-		if p != nil {
-			if content, ok := p.([]byte); ok {
-				history = append(history, content...)
-			}
-		}
-	})
-	return history
+// Clear releases the buffer memory, making it eligible for GC.
+// The buffer will be lazily re-allocated on the next Write.
+func (w *LogMonitor) Clear() {
+	w.bufferMu.Lock()
+	w.buffer = nil
+	w.bufferMu.Unlock()
 }
 
 func (w *LogMonitor) OnLogData(callback func(data []byte)) context.CancelFunc {
@@ -106,12 +196,22 @@ func (w *LogMonitor) SetLogLevel(level LogLevel) {
 	w.level = level
 }
 
+func (w *LogMonitor) SetLogTimeFormat(timeFormat string) {
+	w.mu.Lock()
+	defer w.mu.Unlock()
+	w.timeFormat = timeFormat
+}
+
 func (w *LogMonitor) formatMessage(level string, msg string) []byte {
 	prefix := ""
 	if w.prefix != "" {
 		prefix = fmt.Sprintf("[%s] ", w.prefix)
 	}
-	return []byte(fmt.Sprintf("%s[%s] %s\n", prefix, level, msg))
+	timestamp := ""
+	if w.timeFormat != "" {
+		timestamp = fmt.Sprintf("%s ", time.Now().Format(w.timeFormat))
+	}
+	return []byte(fmt.Sprintf("%s%s[%s] %s\n", timestamp, prefix, level, msg))
 }
 
 func (w *LogMonitor) log(level LogLevel, msg string) {

proxy/logMonitor_test.go

@@ -3,8 +3,10 @@ package proxy
 import (
 	"bytes"
 	"io"
+	"strings"
 	"sync"
 	"testing"
+	"time"
 )
 
 func TestLogMonitor(t *testing.T) {
@@ -84,3 +86,231 @@ func TestWrite_ImmutableBuffer(t *testing.T) {
 		t.Errorf("Expected history to be %q, got %q", expected, history)
 	}
 }
+
+func TestWrite_LogTimeFormat(t *testing.T) {
+	// Create a new LogMonitor instance
+	lm := NewLogMonitorWriter(io.Discard)
+
+	// Enable timestamps
+	lm.timeFormat = time.RFC3339
+
+	// Write the message to the LogMonitor
+	lm.Info("Hello, World!")
+
+	// Get the history from the LogMonitor
+	history := lm.GetHistory()
+
+	timestamp := ""
+	fields := strings.Fields(string(history))
+	if len(fields) > 0 {
+		timestamp = fields[0]
+	} else {
+		t.Fatalf("Cannot extract string from history")
+	}
+
+	_, err := time.Parse(time.RFC3339, timestamp)
+	if err != nil {
+		t.Fatalf("Cannot find timestamp: %v", err)
+	}
+}
+
+func TestCircularBuffer_WrapAround(t *testing.T) {
+	// Create a small buffer to test wrap-around
+	cb := newCircularBuffer(10)
+
+	// Write "hello" (5 bytes)
+	cb.Write([]byte("hello"))
+	if got := string(cb.GetHistory()); got != "hello" {
+		t.Errorf("Expected 'hello', got %q", got)
+	}
+
+	// Write "world" (5 bytes) - buffer now full
+	cb.Write([]byte("world"))
+	if got := string(cb.GetHistory()); got != "helloworld" {
+		t.Errorf("Expected 'helloworld', got %q", got)
+	}
+
+	// Write "12345" (5 bytes) - should overwrite "hello"
+	cb.Write([]byte("12345"))
+	if got := string(cb.GetHistory()); got != "world12345" {
+		t.Errorf("Expected 'world12345', got %q", got)
+	}
+
+	// Write data larger than buffer capacity
+	cb.Write([]byte("abcdefghijklmnop")) // 16 bytes, only last 10 kept
+	if got := string(cb.GetHistory()); got != "ghijklmnop" {
+		t.Errorf("Expected 'ghijklmnop', got %q", got)
+	}
+}
+
+func TestCircularBuffer_BoundaryConditions(t *testing.T) {
+	// Test empty buffer
+	cb := newCircularBuffer(10)
+	if got := cb.GetHistory(); got != nil {
+		t.Errorf("Expected nil for empty buffer, got %q", got)
+	}
+
+	// Test exact capacity
+	cb.Write([]byte("1234567890"))
+	if got := string(cb.GetHistory()); got != "1234567890" {
+		t.Errorf("Expected '1234567890', got %q", got)
+	}
+
+	// Test write exactly at capacity boundary
+	cb = newCircularBuffer(10)
+	cb.Write([]byte("12345"))
+	cb.Write([]byte("67890"))
+	if got := string(cb.GetHistory()); got != "1234567890" {
+		t.Errorf("Expected '1234567890', got %q", got)
+	}
+}
+
+func TestLogMonitor_LazyInit(t *testing.T) {
+	lm := NewLogMonitorWriter(io.Discard)
+
+	// Buffer should be nil before any writes
+	if lm.buffer != nil {
+		t.Error("Expected buffer to be nil before first write")
+	}
+
+	// GetHistory should return nil when buffer is nil
+	if got := lm.GetHistory(); got != nil {
+		t.Errorf("Expected nil history before first write, got %q", got)
+	}
+
+	// Write should lazily initialize the buffer
+	lm.Write([]byte("test"))
+
+	if lm.buffer == nil {
+		t.Error("Expected buffer to be initialized after write")
+	}
+
+	if got := string(lm.GetHistory()); got != "test" {
+		t.Errorf("Expected 'test', got %q", got)
+	}
+}
+
+func TestLogMonitor_Clear(t *testing.T) {
+	lm := NewLogMonitorWriter(io.Discard)
+
+	// Write some data
+	lm.Write([]byte("hello"))
+	if got := string(lm.GetHistory()); got != "hello" {
+		t.Errorf("Expected 'hello', got %q", got)
+	}
+
+	// Clear should release the buffer
+	lm.Clear()
+
+	if lm.buffer != nil {
+		t.Error("Expected buffer to be nil after Clear")
+	}
+
+	if got := lm.GetHistory(); got != nil {
+		t.Errorf("Expected nil history after Clear, got %q", got)
+	}
+}
+
+func TestLogMonitor_ClearAndReuse(t *testing.T) {
+	lm := NewLogMonitorWriter(io.Discard)
+
+	// Write, clear, then write again
+	lm.Write([]byte("first"))
+	lm.Clear()
+	lm.Write([]byte("second"))
+
+	if got := string(lm.GetHistory()); got != "second" {
+		t.Errorf("Expected 'second' after clear and reuse, got %q", got)
+	}
+}
+
+func BenchmarkLogMonitorWrite(b *testing.B) {
+	// Test data of varying sizes
+	smallMsg := []byte("small message\n")
+	mediumMsg := []byte(strings.Repeat("medium message content ", 10) + "\n")
+	largeMsg := []byte(strings.Repeat("large message content for benchmarking ", 100) + "\n")
+
+	b.Run("SmallWrite", func(b *testing.B) {
+		lm := NewLogMonitorWriter(io.Discard)
+		b.ResetTimer()
+		for i := 0; i < b.N; i++ {
+			lm.Write(smallMsg)
+		}
+	})
+
+	b.Run("MediumWrite", func(b *testing.B) {
+		lm := NewLogMonitorWriter(io.Discard)
+		b.ResetTimer()
+		for i := 0; i < b.N; i++ {
+			lm.Write(mediumMsg)
+		}
+	})
+
+	b.Run("LargeWrite", func(b *testing.B) {
+		lm := NewLogMonitorWriter(io.Discard)
+		b.ResetTimer()
+		for i := 0; i < b.N; i++ {
+			lm.Write(largeMsg)
+		}
+	})
+
+	b.Run("WithSubscribers", func(b *testing.B) {
+		lm := NewLogMonitorWriter(io.Discard)
+		// Add some subscribers
+		for i := 0; i < 5; i++ {
+			lm.OnLogData(func(data []byte) {})
+		}
+		b.ResetTimer()
+		for i := 0; i < b.N; i++ {
+			lm.Write(mediumMsg)
+		}
+	})
+
+	b.Run("GetHistory", func(b *testing.B) {
+		lm := NewLogMonitorWriter(io.Discard)
+		// Pre-populate with data
+		for i := 0; i < 1000; i++ {
+			lm.Write(mediumMsg)
+		}
+		b.ResetTimer()
+		for i := 0; i < b.N; i++ {
+			lm.GetHistory()
+		}
+	})
+}
+
+/*
+Benchmark Results - MBP M1 Pro
+
+Before (ring.Ring):
+| Benchmark                       | ns/op      | bytes/op | allocs/op |
+|---------------------------------|------------|----------|-----------|
+| SmallWrite (14B)                | 43 ns      | 40 B     | 2         |
+| MediumWrite (241B)              | 76 ns      | 264 B    | 2         |
+| LargeWrite (4KB)                | 504 ns     | 4,120 B  | 2         |
+| WithSubscribers (5 subs)        | 355 ns     | 264 B    | 2         |
+| GetHistory (after 1000 writes)  | 145,000 ns | 1.2 MB   | 22        |
+
+After (circularBuffer 10KB):
+| Benchmark                       | ns/op      | bytes/op | allocs/op |
+|---------------------------------|------------|----------|-----------|
+| SmallWrite (14B)                | 26 ns      | 16 B     | 1         |
+| MediumWrite (241B)              | 67 ns      | 240 B    | 1         |
+| LargeWrite (4KB)                | 774 ns     | 4,096 B  | 1         |
+| WithSubscribers (5 subs)        | 325 ns     | 240 B    | 1         |
+| GetHistory (after 1000 writes)  | 1,042 ns   | 10,240 B | 1         |
+
+After (circularBuffer 100KB):
+| Benchmark                       | ns/op      | bytes/op  | allocs/op |
+|---------------------------------|------------|-----------|-----------|
+| SmallWrite (14B)                | 26 ns      | 16 B      | 1         |
+| MediumWrite (241B)              | 66 ns      | 240 B     | 1         |
+| LargeWrite (4KB)                | 753 ns     | 4,096 B   | 1         |
+| WithSubscribers (5 subs)        | 309 ns     | 240 B     | 1         |
+| GetHistory (after 1000 writes)  | 7,788 ns   | 106,496 B | 1         |
+
+Summary:
+- GetHistory: 139x faster (10KB), 18x faster (100KB)
+- Allocations: reduced from 2 to 1 across all operations
+- Small/medium writes: ~1.1-1.6x faster
+*/

proxy/metrics_middleware.go

@@ -1,184 +0,0 @@
-package proxy
-
-import (
-	"bytes"
-	"fmt"
-	"io"
-	"net/http"
-	"strings"
-	"time"
-
-	"github.com/gin-gonic/gin"
-	"github.com/tidwall/gjson"
-)
-
-type MetricsRecorder struct {
-	metricsMonitor *MetricsMonitor
-	realModelName  string
-	//	isStreaming    bool
-	startTime time.Time
-}
-
-// MetricsMiddleware sets up the MetricsResponseWriter for capturing upstream requests
-func MetricsMiddleware(pm *ProxyManager) gin.HandlerFunc {
-	return func(c *gin.Context) {
-		bodyBytes, err := io.ReadAll(c.Request.Body)
-		if err != nil {
-			pm.sendErrorResponse(c, http.StatusBadRequest, "could not ready request body")
-			c.Abort()
-			return
-		}
-		c.Request.Body = io.NopCloser(bytes.NewBuffer(bodyBytes))
-
-		requestedModel := gjson.GetBytes(bodyBytes, "model").String()
-		if requestedModel == "" {
-			pm.sendErrorResponse(c, http.StatusBadRequest, "missing or invalid 'model' key")
-			c.Abort()
-			return
-		}
-
-		realModelName, found := pm.config.RealModelName(requestedModel)
-		if !found {
-			pm.sendErrorResponse(c, http.StatusBadRequest, fmt.Sprintf("could not find real modelID for %s", requestedModel))
-			c.Abort()
-			return
-		}
-
-		writer := &MetricsResponseWriter{
-			ResponseWriter: c.Writer,
-			metricsRecorder: &MetricsRecorder{
-				metricsMonitor: pm.metricsMonitor,
-				realModelName:  realModelName,
-				startTime:      time.Now(),
-			},
-		}
-		c.Writer = writer
-		c.Next()
-
-		// check for streaming response
-		if strings.Contains(c.Writer.Header().Get("Content-Type"), "text/event-stream") {
-			writer.metricsRecorder.processStreamingResponse(writer.body)
-		} else {
-			writer.metricsRecorder.processNonStreamingResponse(writer.body)
-		}
-	}
-}
-
-func (rec *MetricsRecorder) parseAndRecordMetrics(jsonData gjson.Result) bool {
-	usage := jsonData.Get("usage")
-	timings := jsonData.Get("timings")
-	if !usage.Exists() && !timings.Exists() {
-		return false
-	}
-
-	// default values
-	cachedTokens := -1 // unknown or missing data
-	outputTokens := 0
-	inputTokens := 0
-
-	// timings data
-	tokensPerSecond := -1.0
-	promptPerSecond := -1.0
-	durationMs := int(time.Since(rec.startTime).Milliseconds())
-
-	if usage.Exists() {
-		outputTokens = int(jsonData.Get("usage.completion_tokens").Int())
-		inputTokens = int(jsonData.Get("usage.prompt_tokens").Int())
-	}
-
-	// use llama-server's timing data for tok/sec and duration as it is more accurate
-	if timings.Exists() {
-		inputTokens = int(jsonData.Get("timings.prompt_n").Int())
-		outputTokens = int(jsonData.Get("timings.predicted_n").Int())
-		promptPerSecond = jsonData.Get("timings.prompt_per_second").Float()
-		tokensPerSecond = jsonData.Get("timings.predicted_per_second").Float()
-		durationMs = int(jsonData.Get("timings.prompt_ms").Float() + jsonData.Get("timings.predicted_ms").Float())
-
-		if cachedValue := jsonData.Get("timings.cache_n"); cachedValue.Exists() {
-			cachedTokens = int(cachedValue.Int())
-		}
-	}
-
-	rec.metricsMonitor.addMetrics(TokenMetrics{
-		Timestamp:       time.Now(),
-		Model:           rec.realModelName,
-		CachedTokens:    cachedTokens,
-		InputTokens:     inputTokens,
-		OutputTokens:    outputTokens,
-		PromptPerSecond: promptPerSecond,
-		TokensPerSecond: tokensPerSecond,
-		DurationMs:      durationMs,
-	})
-
-	return true
-}
-
-func (rec *MetricsRecorder) processStreamingResponse(body []byte) {
-	// Iterate **backwards** through the lines looking for the data payload with
-	// usage data
-	lines := bytes.Split(body, []byte("\n"))
-
-	for i := len(lines) - 1; i >= 0; i-- {
-		line := bytes.TrimSpace(lines[i])
-		if len(line) == 0 {
-			continue
-		}
-
-		// SSE payload always follows "data:"
-		prefix := []byte("data:")
-		if !bytes.HasPrefix(line, prefix) {
-			continue
-		}
-		data := bytes.TrimSpace(line[len(prefix):])
-
-		if len(data) == 0 {
-			continue
-		}
-
-		if bytes.Equal(data, []byte("[DONE]")) {
-			// [DONE] line itself contains nothing of interest.
-			continue
-		}
-
-		if gjson.ValidBytes(data) {
-			if rec.parseAndRecordMetrics(gjson.ParseBytes(data)) {
-				return // short circuit if a metric was recorded
-			}
-		}
-	}
-}
-
-func (rec *MetricsRecorder) processNonStreamingResponse(body []byte) {
-	if len(body) == 0 {
-		return
-	}
-
-	// Parse JSON to extract usage information
-	if gjson.ValidBytes(body) {
-		rec.parseAndRecordMetrics(gjson.ParseBytes(body))
-	}
-}
-
-// MetricsResponseWriter captures the entire response for non-streaming
-type MetricsResponseWriter struct {
-	gin.ResponseWriter
-	body            []byte
-	metricsRecorder *MetricsRecorder
-}
-
-func (w *MetricsResponseWriter) Write(b []byte) (int, error) {
-	n, err := w.ResponseWriter.Write(b)
-	if err != nil {
-		return n, err
-	}
-	w.body = append(w.body, b...)
-	return n, nil
-}
-
-func (w *MetricsResponseWriter) WriteHeader(statusCode int) {
-	w.ResponseWriter.WriteHeader(statusCode)
-}
-
-func (w *MetricsResponseWriter) Header() http.Header {
-	return w.ResponseWriter.Header()
-}

proxy/metrics_monitor.go

@@ -1,12 +1,18 @@
 package proxy
 
 import (
+	"bytes"
 	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"strings"
 	"sync"
 	"time"
 
+	"github.com/gin-gonic/gin"
 	"github.com/mostlygeek/llama-swap/event"
-	"github.com/mostlygeek/llama-swap/proxy/config"
+	"github.com/tidwall/gjson"
 )
 
 // TokenMetrics represents parsed token statistics from llama-server logs
@@ -31,21 +37,18 @@ func (e TokenMetricsEvent) Type() uint32 {
 	return TokenMetricsEventID // defined in events.go
 }
 
-// MetricsMonitor parses llama-server output for token statistics
-type MetricsMonitor struct {
+// metricsMonitor parses llama-server output for token statistics
+type metricsMonitor struct {
 	mu         sync.RWMutex
 	metrics    []TokenMetrics
 	maxMetrics int
 	nextID     int
+	logger     *LogMonitor
 }
 
-func NewMetricsMonitor(config *config.Config) *MetricsMonitor {
-	maxMetrics := config.MetricsMaxInMemory
-	if maxMetrics <= 0 {
-		maxMetrics = 1000 // Default fallback
-	}
-
-	mp := &MetricsMonitor{
+func newMetricsMonitor(logger *LogMonitor, maxMetrics int) *metricsMonitor {
+	mp := &metricsMonitor{
+		logger:     logger,
 		maxMetrics: maxMetrics,
 	}
 
@@ -53,7 +56,7 @@ func NewMetricsMonitor(config *config.Config) *MetricsMonitor {
 }
 
 // addMetrics adds a new metric to the collection and publishes an event
-func (mp *MetricsMonitor) addMetrics(metric TokenMetrics) {
+func (mp *metricsMonitor) addMetrics(metric TokenMetrics) {
 	mp.mu.Lock()
 	defer mp.mu.Unlock()
 
@@ -66,8 +69,8 @@ func (mp *MetricsMonitor) addMetrics(metric TokenMetrics) {
 	event.Emit(TokenMetricsEvent{Metrics: metric})
 }
 
-// GetMetrics returns a copy of the current metrics
-func (mp *MetricsMonitor) GetMetrics() []TokenMetrics {
+// getMetrics returns a copy of the current metrics
+func (mp *metricsMonitor) getMetrics() []TokenMetrics {
 	mp.mu.RLock()
 	defer mp.mu.RUnlock()
 
@@ -76,9 +79,213 @@ func (mp *MetricsMonitor) GetMetrics() []TokenMetrics {
 	return result
 }
 
-// GetMetricsJSON returns metrics as JSON
-func (mp *MetricsMonitor) GetMetricsJSON() ([]byte, error) {
+// getMetricsJSON returns metrics as JSON
+func (mp *metricsMonitor) getMetricsJSON() ([]byte, error) {
 	mp.mu.RLock()
 	defer mp.mu.RUnlock()
 	return json.Marshal(mp.metrics)
 }
+
+// wrapHandler wraps the proxy handler to extract token metrics
+// if wrapHandler returns an error it is safe to assume that no
+// data was sent to the client
+func (mp *metricsMonitor) wrapHandler(
+	modelID string,
+	writer gin.ResponseWriter,
+	request *http.Request,
+	next func(modelID string, w http.ResponseWriter, r *http.Request) error,
+) error {
+	recorder := newBodyCopier(writer)
+	if err := next(modelID, recorder, request); err != nil {
+		return err
+	}
+
+	// after this point we have to assume that data was sent to the client
+	// and we can only log errors but not send them to clients
+
+	if recorder.Status() != http.StatusOK {
+		mp.logger.Warnf("metrics skipped, HTTP status=%d, path=%s", recorder.Status(), request.URL.Path)
+		return nil
+	}
+
+	body := recorder.body.Bytes()
+	if len(body) == 0 {
+		mp.logger.Warn("metrics skipped, empty body")
+		return nil
+	}
+
+	if strings.Contains(recorder.Header().Get("Content-Type"), "text/event-stream") {
+		if tm, err := processStreamingResponse(modelID, recorder.StartTime(), body); err != nil {
+			mp.logger.Warnf("error processing streaming response: %v, path=%s", err, request.URL.Path)
+		} else {
+			mp.addMetrics(tm)
+		}
+	} else {
+		if gjson.ValidBytes(body) {
+			parsed := gjson.ParseBytes(body)
+			usage := parsed.Get("usage")
+			timings := parsed.Get("timings")
+
+			if usage.Exists() || timings.Exists() {
+				if tm, err := parseMetrics(modelID, recorder.StartTime(), usage, timings); err != nil {
+					mp.logger.Warnf("error parsing metrics: %v, path=%s", err, request.URL.Path)
+				} else {
+					mp.addMetrics(tm)
+				}
+			}
+
+		} else {
+			mp.logger.Warnf("metrics skipped, invalid JSON in response body path=%s", request.URL.Path)
+		}
+	}
+
+	return nil
+}
+
+func processStreamingResponse(modelID string, start time.Time, body []byte) (TokenMetrics, error) {
+	// Iterate **backwards** through the body looking for the data payload with
+	// usage data. This avoids allocating a slice of all lines via bytes.Split.
+
+	// Start from the end of the body and scan backwards for newlines
+	pos := len(body)
+	for pos > 0 {
+		// Find the previous newline (or start of body)
+		lineStart := bytes.LastIndexByte(body[:pos], '\n')
+		if lineStart == -1 {
+			lineStart = 0
+		} else {
+			lineStart++ // Move past the newline
+		}
+
+		line := bytes.TrimSpace(body[lineStart:pos])
+		pos = lineStart - 1 // Move position before the newline for next iteration
+
+		if len(line) == 0 {
+			continue
+		}
+
+		// SSE payload always follows "data:"
+		prefix := []byte("data:")
+		if !bytes.HasPrefix(line, prefix) {
+			continue
+		}
+		data := bytes.TrimSpace(line[len(prefix):])
+
+		if len(data) == 0 {
+			continue
+		}
+
+		if bytes.Equal(data, []byte("[DONE]")) {
+			// [DONE] line itself contains nothing of interest.
+			continue
+		}
+
+		if gjson.ValidBytes(data) {
+			parsed := gjson.ParseBytes(data)
+			usage := parsed.Get("usage")
+			timings := parsed.Get("timings")
+
+			if usage.Exists() || timings.Exists() {
+				return parseMetrics(modelID, start, usage, timings)
+			}
+		}
+	}
+
+	return TokenMetrics{}, fmt.Errorf("no valid JSON data found in stream")
+}
+
+func parseMetrics(modelID string, start time.Time, usage, timings gjson.Result) (TokenMetrics, error) {
+	// default values
+	cachedTokens := -1 // unknown or missing data
+	outputTokens := 0
+	inputTokens := 0
+
+	// timings data
+	tokensPerSecond := -1.0
+	promptPerSecond := -1.0
+	durationMs := int(time.Since(start).Milliseconds())
+
+	if usage.Exists() {
+		if pt := usage.Get("prompt_tokens"); pt.Exists() {
+			// v1/chat/completions
+			inputTokens = int(pt.Int())
+		} else if it := usage.Get("input_tokens"); it.Exists() {
+			// v1/messages
+			inputTokens = int(it.Int())
+		}
+
+		if ct := usage.Get("completion_tokens"); ct.Exists() {
+			// v1/chat/completions
+			outputTokens = int(ct.Int())
+		} else if ot := usage.Get("output_tokens"); ot.Exists() {
+			outputTokens = int(ot.Int())
+		}
+
+		if ct := usage.Get("cache_read_input_tokens"); ct.Exists() {
+			cachedTokens = int(ct.Int())
+		}
+	}
+
+	// use llama-server's timing data for tok/sec and duration as it is more accurate
+	if timings.Exists() {
+		inputTokens = int(timings.Get("prompt_n").Int())
+		outputTokens = int(timings.Get("predicted_n").Int())
+		promptPerSecond = timings.Get("prompt_per_second").Float()
+		tokensPerSecond = timings.Get("predicted_per_second").Float()
+		durationMs = int(timings.Get("prompt_ms").Float() + timings.Get("predicted_ms").Float())
+
+		if cachedValue := timings.Get("cache_n"); cachedValue.Exists() {
+			cachedTokens = int(cachedValue.Int())
+		}
+	}
+
+	return TokenMetrics{
+		Timestamp:       time.Now(),
+		Model:           modelID,
+		CachedTokens:    cachedTokens,
+		InputTokens:     inputTokens,
+		OutputTokens:    outputTokens,
+		PromptPerSecond: promptPerSecond,
+		TokensPerSecond: tokensPerSecond,
+		DurationMs:      durationMs,
+	}, nil
+}
+
+// responseBodyCopier records the response body and writes to the original response writer
+// while also capturing it in a buffer for later processing
+type responseBodyCopier struct {
+	gin.ResponseWriter
+	body  *bytes.Buffer
+	tee   io.Writer
+	start time.Time
+}
+
+func newBodyCopier(w gin.ResponseWriter) *responseBodyCopier {
+	bodyBuffer := &bytes.Buffer{}
+	return &responseBodyCopier{
+		ResponseWriter: w,
+		body:           bodyBuffer,
+		tee:            io.MultiWriter(w, bodyBuffer),
+	}
+}
+
+func (w *responseBodyCopier) Write(b []byte) (int, error) {
+	if w.start.IsZero() {
+		w.start = time.Now()
+	}
+
+	// Single write operation that writes to both the response and buffer
+	return w.tee.Write(b)
+}
+
+func (w *responseBodyCopier) WriteHeader(statusCode int) {
+	w.ResponseWriter.WriteHeader(statusCode)
+}
+
+func (w *responseBodyCopier) Header() http.Header {
+	return w.ResponseWriter.Header()
+}
+
+func (w *responseBodyCopier) StartTime() time.Time {
+	return w.start
+}

proxy/metrics_monitor_test.go

@@ -0,0 +1,693 @@
+package proxy
+
+import (
+	"encoding/json"
+	"net/http"
+	"net/http/httptest"
+	"sync"
+	"testing"
+	"time"
+
+	"github.com/gin-gonic/gin"
+	"github.com/mostlygeek/llama-swap/event"
+	"github.com/stretchr/testify/assert"
+)
+
+func TestMetricsMonitor_AddMetrics(t *testing.T) {
+	t.Run("adds metrics and assigns ID", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		metric := TokenMetrics{
+			Model:        "test-model",
+			InputTokens:  100,
+			OutputTokens: 50,
+		}
+
+		mm.addMetrics(metric)
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 1, len(metrics))
+		assert.Equal(t, 0, metrics[0].ID)
+		assert.Equal(t, "test-model", metrics[0].Model)
+		assert.Equal(t, 100, metrics[0].InputTokens)
+		assert.Equal(t, 50, metrics[0].OutputTokens)
+	})
+
+	t.Run("increments ID for each metric", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		for i := 0; i < 5; i++ {
+			mm.addMetrics(TokenMetrics{Model: "model"})
+		}
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 5, len(metrics))
+		for i := 0; i < 5; i++ {
+			assert.Equal(t, i, metrics[i].ID)
+		}
+	})
+
+	t.Run("respects max metrics limit", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 3)
+
+		// Add 5 metrics
+		for i := 0; i < 5; i++ {
+			mm.addMetrics(TokenMetrics{
+				Model:       "model",
+				InputTokens: i,
+			})
+		}
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 3, len(metrics))
+
+		// Should keep the last 3 metrics (IDs 2, 3, 4)
+		assert.Equal(t, 2, metrics[0].ID)
+		assert.Equal(t, 3, metrics[1].ID)
+		assert.Equal(t, 4, metrics[2].ID)
+	})
+
+	t.Run("emits TokenMetricsEvent", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		receivedEvent := make(chan TokenMetricsEvent, 1)
+		cancel := event.On(func(e TokenMetricsEvent) {
+			receivedEvent <- e
+		})
+		defer cancel()
+
+		metric := TokenMetrics{
+			Model:        "test-model",
+			InputTokens:  100,
+			OutputTokens: 50,
+		}
+
+		mm.addMetrics(metric)
+
+		select {
+		case evt := <-receivedEvent:
+			assert.Equal(t, 0, evt.Metrics.ID)
+			assert.Equal(t, "test-model", evt.Metrics.Model)
+			assert.Equal(t, 100, evt.Metrics.InputTokens)
+			assert.Equal(t, 50, evt.Metrics.OutputTokens)
+		case <-time.After(1 * time.Second):
+			t.Fatal("timeout waiting for event")
+		}
+	})
+}
+
+func TestMetricsMonitor_GetMetrics(t *testing.T) {
+	t.Run("returns empty slice when no metrics", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+		metrics := mm.getMetrics()
+		assert.NotNil(t, metrics)
+		assert.Equal(t, 0, len(metrics))
+	})
+
+	t.Run("returns copy of metrics", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+		mm.addMetrics(TokenMetrics{Model: "model1"})
+		mm.addMetrics(TokenMetrics{Model: "model2"})
+
+		metrics1 := mm.getMetrics()
+		metrics2 := mm.getMetrics()
+
+		// Verify we got copies
+		assert.Equal(t, 2, len(metrics1))
+		assert.Equal(t, 2, len(metrics2))
+
+		// Modify the returned slice shouldn't affect the original
+		metrics1[0].Model = "modified"
+		metrics3 := mm.getMetrics()
+		assert.Equal(t, "model1", metrics3[0].Model)
+	})
+}
+
+func TestMetricsMonitor_GetMetricsJSON(t *testing.T) {
+	t.Run("returns valid JSON for empty metrics", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+		jsonData, err := mm.getMetricsJSON()
+		assert.NoError(t, err)
+		assert.NotNil(t, jsonData)
+
+		var metrics []TokenMetrics
+		err = json.Unmarshal(jsonData, &metrics)
+		assert.NoError(t, err)
+		assert.Equal(t, 0, len(metrics))
+	})
+
+	t.Run("returns valid JSON with metrics", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+		mm.addMetrics(TokenMetrics{
+			Model:           "model1",
+			InputTokens:     100,
+			OutputTokens:    50,
+			TokensPerSecond: 25.5,
+		})
+		mm.addMetrics(TokenMetrics{
+			Model:           "model2",
+			InputTokens:     200,
+			OutputTokens:    100,
+			TokensPerSecond: 30.0,
+		})
+
+		jsonData, err := mm.getMetricsJSON()
+		assert.NoError(t, err)
+
+		var metrics []TokenMetrics
+		err = json.Unmarshal(jsonData, &metrics)
+		assert.NoError(t, err)
+		assert.Equal(t, 2, len(metrics))
+		assert.Equal(t, "model1", metrics[0].Model)
+		assert.Equal(t, "model2", metrics[1].Model)
+	})
+}
+
+func TestMetricsMonitor_WrapHandler(t *testing.T) {
+	t.Run("successful non-streaming request with usage data", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		responseBody := `{
+			"usage": {
+				"prompt_tokens": 100,
+				"completion_tokens": 50
+			}
+		}`
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.Header().Set("Content-Type", "application/json")
+			w.WriteHeader(http.StatusOK)
+			w.Write([]byte(responseBody))
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.NoError(t, err)
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 1, len(metrics))
+		assert.Equal(t, "test-model", metrics[0].Model)
+		assert.Equal(t, 100, metrics[0].InputTokens)
+		assert.Equal(t, 50, metrics[0].OutputTokens)
+	})
+
+	t.Run("successful request with timings data", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		responseBody := `{
+			"timings": {
+				"prompt_n": 100,
+				"predicted_n": 50,
+				"prompt_per_second": 150.5,
+				"predicted_per_second": 25.5,
+				"prompt_ms": 500.0,
+				"predicted_ms": 1500.0,
+				"cache_n": 20
+			}
+		}`
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.Header().Set("Content-Type", "application/json")
+			w.WriteHeader(http.StatusOK)
+			w.Write([]byte(responseBody))
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.NoError(t, err)
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 1, len(metrics))
+		assert.Equal(t, "test-model", metrics[0].Model)
+		assert.Equal(t, 100, metrics[0].InputTokens)
+		assert.Equal(t, 50, metrics[0].OutputTokens)
+		assert.Equal(t, 20, metrics[0].CachedTokens)
+		assert.Equal(t, 150.5, metrics[0].PromptPerSecond)
+		assert.Equal(t, 25.5, metrics[0].TokensPerSecond)
+		assert.Equal(t, 2000, metrics[0].DurationMs) // 500 + 1500
+	})
+
+	t.Run("streaming request with SSE format", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		// Note: SSE format requires proper line breaks - each data line followed by blank line
+		responseBody := `data: {"choices":[{"text":"Hello"}]}
+
+data: {"choices":[{"text":" World"}]}
+
+data: {"usage":{"prompt_tokens":10,"completion_tokens":20},"timings":{"prompt_n":10,"predicted_n":20,"prompt_per_second":100.0,"predicted_per_second":50.0,"prompt_ms":100.0,"predicted_ms":400.0}}
+
+data: [DONE]
+
+`
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.Header().Set("Content-Type", "text/event-stream")
+			w.WriteHeader(http.StatusOK)
+			w.Write([]byte(responseBody))
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.NoError(t, err)
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 1, len(metrics))
+		assert.Equal(t, "test-model", metrics[0].Model)
+		// When timings data is present, it takes precedence
+		assert.Equal(t, 10, metrics[0].InputTokens)
+		assert.Equal(t, 20, metrics[0].OutputTokens)
+	})
+
+	t.Run("non-OK status code does not record metrics", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.WriteHeader(http.StatusBadRequest)
+			w.Write([]byte("error"))
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.NoError(t, err)
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 0, len(metrics))
+	})
+
+	t.Run("empty response body does not record metrics", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.WriteHeader(http.StatusOK)
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.NoError(t, err)
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 0, len(metrics))
+	})
+
+	t.Run("invalid JSON does not record metrics", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.Header().Set("Content-Type", "application/json")
+			w.WriteHeader(http.StatusOK)
+			w.Write([]byte("not valid json"))
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.NoError(t, err) // Errors after response is sent are logged, not returned
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 0, len(metrics))
+	})
+
+	t.Run("next handler error is propagated", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		expectedErr := assert.AnError
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			return expectedErr
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.Equal(t, expectedErr, err)
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 0, len(metrics))
+	})
+
+	t.Run("response without usage or timings does not record metrics", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		responseBody := `{"result": "ok"}`
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.Header().Set("Content-Type", "application/json")
+			w.WriteHeader(http.StatusOK)
+			w.Write([]byte(responseBody))
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.NoError(t, err) // Errors after response is sent are logged, not returned
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 0, len(metrics))
+	})
+}
+
+func TestMetricsMonitor_ResponseBodyCopier(t *testing.T) {
+	t.Run("captures response body", func(t *testing.T) {
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+		copier := newBodyCopier(ginCtx.Writer)
+
+		testData := []byte("test response body")
+		n, err := copier.Write(testData)
+
+		assert.NoError(t, err)
+		assert.Equal(t, len(testData), n)
+		assert.Equal(t, testData, copier.body.Bytes())
+		assert.Equal(t, string(testData), rec.Body.String())
+	})
+
+	t.Run("sets start time on first write", func(t *testing.T) {
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+		copier := newBodyCopier(ginCtx.Writer)
+
+		assert.True(t, copier.StartTime().IsZero())
+
+		copier.Write([]byte("test"))
+
+		assert.False(t, copier.StartTime().IsZero())
+	})
+
+	t.Run("preserves headers", func(t *testing.T) {
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+		copier := newBodyCopier(ginCtx.Writer)
+
+		copier.Header().Set("X-Test", "value")
+
+		assert.Equal(t, "value", rec.Header().Get("X-Test"))
+	})
+
+	t.Run("preserves status code", func(t *testing.T) {
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+		copier := newBodyCopier(ginCtx.Writer)
+
+		copier.WriteHeader(http.StatusCreated)
+
+		// Gin's ResponseWriter tracks status internally
+		assert.Equal(t, http.StatusCreated, copier.Status())
+	})
+}
+
+func TestMetricsMonitor_Concurrent(t *testing.T) {
+	t.Run("concurrent addMetrics is safe", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 1000)
+
+		var wg sync.WaitGroup
+		numGoroutines := 10
+		metricsPerGoroutine := 100
+
+		for i := 0; i < numGoroutines; i++ {
+			wg.Add(1)
+			go func(id int) {
+				defer wg.Done()
+				for j := 0; j < metricsPerGoroutine; j++ {
+					mm.addMetrics(TokenMetrics{
+						Model:        "test-model",
+						InputTokens:  id*1000 + j,
+						OutputTokens: j,
+					})
+				}
+			}(i)
+		}
+
+		wg.Wait()
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, numGoroutines*metricsPerGoroutine, len(metrics))
+	})
+
+	t.Run("concurrent reads and writes are safe", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 100)
+
+		done := make(chan bool)
+
+		// Writer goroutine
+		go func() {
+			for i := 0; i < 50; i++ {
+				mm.addMetrics(TokenMetrics{Model: "test-model"})
+				time.Sleep(1 * time.Millisecond)
+			}
+			done <- true
+		}()
+
+		// Multiple reader goroutines
+		var wg sync.WaitGroup
+		for i := 0; i < 5; i++ {
+			wg.Add(1)
+			go func() {
+				defer wg.Done()
+				for j := 0; j < 20; j++ {
+					_ = mm.getMetrics()
+					_, _ = mm.getMetricsJSON()
+					time.Sleep(2 * time.Millisecond)
+				}
+			}()
+		}
+
+		<-done
+		wg.Wait()
+
+		// Final check
+		metrics := mm.getMetrics()
+		assert.Equal(t, 50, len(metrics))
+	})
+}
+
+func TestMetricsMonitor_ParseMetrics(t *testing.T) {
+	t.Run("prefers timings over usage data", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		// Timings should take precedence over usage
+		responseBody := `{
+			"usage": {
+				"prompt_tokens": 50,
+				"completion_tokens": 25
+			},
+			"timings": {
+				"prompt_n": 100,
+				"predicted_n": 50,
+				"prompt_per_second": 150.5,
+				"predicted_per_second": 25.5,
+				"prompt_ms": 500.0,
+				"predicted_ms": 1500.0
+			}
+		}`
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.Header().Set("Content-Type", "application/json")
+			w.WriteHeader(http.StatusOK)
+			w.Write([]byte(responseBody))
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.NoError(t, err)
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 1, len(metrics))
+		// Should use timings values, not usage values
+		assert.Equal(t, 100, metrics[0].InputTokens)
+		assert.Equal(t, 50, metrics[0].OutputTokens)
+	})
+
+	t.Run("handles missing cache_n in timings", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		responseBody := `{
+			"timings": {
+				"prompt_n": 100,
+				"predicted_n": 50,
+				"prompt_per_second": 150.5,
+				"predicted_per_second": 25.5,
+				"prompt_ms": 500.0,
+				"predicted_ms": 1500.0
+			}
+		}`
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.Header().Set("Content-Type", "application/json")
+			w.WriteHeader(http.StatusOK)
+			w.Write([]byte(responseBody))
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.NoError(t, err)
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 1, len(metrics))
+		assert.Equal(t, -1, metrics[0].CachedTokens) // Default value when not present
+	})
+}
+
+func TestMetricsMonitor_StreamingResponse(t *testing.T) {
+	t.Run("finds metrics in last valid SSE data", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		// Metrics should be found in the last data line before [DONE]
+		responseBody := `data: {"choices":[{"text":"First"}]}
+
+data: {"choices":[{"text":"Second"}]}
+
+data: {"usage":{"prompt_tokens":100,"completion_tokens":50}}
+
+data: [DONE]
+
+`
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.Header().Set("Content-Type", "text/event-stream")
+			w.WriteHeader(http.StatusOK)
+			w.Write([]byte(responseBody))
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.NoError(t, err)
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 1, len(metrics))
+		assert.Equal(t, 100, metrics[0].InputTokens)
+		assert.Equal(t, 50, metrics[0].OutputTokens)
+	})
+
+	t.Run("handles streaming with no valid JSON", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		responseBody := `data: not json
+
+data: [DONE]
+
+`
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.Header().Set("Content-Type", "text/event-stream")
+			w.WriteHeader(http.StatusOK)
+			w.Write([]byte(responseBody))
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		assert.NoError(t, err) // Errors after response is sent are logged, not returned
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 0, len(metrics))
+	})
+
+	t.Run("handles empty streaming response", func(t *testing.T) {
+		mm := newMetricsMonitor(testLogger, 10)
+
+		responseBody := ``
+
+		nextHandler := func(modelID string, w http.ResponseWriter, r *http.Request) error {
+			w.Header().Set("Content-Type", "text/event-stream")
+			w.WriteHeader(http.StatusOK)
+			w.Write([]byte(responseBody))
+			return nil
+		}
+
+		req := httptest.NewRequest("POST", "/test", nil)
+		rec := httptest.NewRecorder()
+		ginCtx, _ := gin.CreateTestContext(rec)
+
+		err := mm.wrapHandler("test-model", ginCtx.Writer, req, nextHandler)
+		// Empty body should not trigger WrapHandler processing
+		assert.NoError(t, err)
+
+		metrics := mm.getMetrics()
+		assert.Equal(t, 0, len(metrics))
+	})
+}
+
+// Benchmark tests
+func BenchmarkMetricsMonitor_AddMetrics(b *testing.B) {
+	mm := newMetricsMonitor(testLogger, 1000)
+
+	metric := TokenMetrics{
+		Model:           "test-model",
+		CachedTokens:    100,
+		InputTokens:     500,
+		OutputTokens:    250,
+		PromptPerSecond: 1200.5,
+		TokensPerSecond: 45.8,
+		DurationMs:      5000,
+		Timestamp:       time.Now(),
+	}
+
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		mm.addMetrics(metric)
+	}
+}
+
+func BenchmarkMetricsMonitor_AddMetrics_SmallBuffer(b *testing.B) {
+	// Test performance with a smaller buffer where wrapping occurs more frequently
+	mm := newMetricsMonitor(testLogger, 100)
+
+	metric := TokenMetrics{
+		Model:           "test-model",
+		CachedTokens:    100,
+		InputTokens:     500,
+		OutputTokens:    250,
+		PromptPerSecond: 1200.5,
+		TokensPerSecond: 45.8,
+		DurationMs:      5000,
+		Timestamp:       time.Now(),
+	}
+
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		mm.addMetrics(metric)
+	}
+}

proxy/process.go

@@ -2,8 +2,10 @@ package proxy
 
 import (
 	"context"
+	"encoding/json"
 	"errors"
 	"fmt"
+	"math/rand"
 	"net"
 	"net/http"
 	"net/http/httputil"
@@ -254,6 +256,7 @@ func (p *Process) start() error {
 	p.cmd.Env = append(p.cmd.Environ(), p.config.Env...)
 	p.cmd.Cancel = p.cmdStopUpstreamProcess
 	p.cmd.WaitDelay = p.gracefulStopTimeout
+	setProcAttributes(p.cmd)
 
 	p.cmdMutex.Lock()
 	p.cancelUpstream = ctxCancelUpstream
@@ -411,6 +414,9 @@ func (p *Process) stopCommand() {
 	stopStartTime := time.Now()
 	defer func() {
 		p.proxyLogger.Debugf("<%s> stopCommand took %v", p.ID, time.Since(stopStartTime))
+
+		// free the buffer in processLogger so the memory can be recovered
+		p.processLogger.Clear()
 	}()
 
 	p.cmdMutex.RLock()
@@ -462,6 +468,12 @@ func (p *Process) checkHealthEndpoint(healthURL string) error {
 }
 
 func (p *Process) ProxyRequest(w http.ResponseWriter, r *http.Request) {
+
+	if p.reverseProxy == nil {
+		http.Error(w, fmt.Sprintf("No reverse proxy available for %s", p.ID), http.StatusInternalServerError)
+		return
+	}
+
 	requestBeginTime := time.Now()
 	var startDuration time.Duration
 
@@ -488,21 +500,68 @@ func (p *Process) ProxyRequest(w http.ResponseWriter, r *http.Request) {
 		p.inFlightRequests.Done()
 	}()
 
+	// for #366
+	// - extract streaming param from request context, should have been set by proxymanager
+	var srw *statusResponseWriter
+	swapCtx, cancelLoadCtx := context.WithCancel(r.Context())
 	// start the process on demand
 	if p.CurrentState() != StateReady {
+		// start a goroutine to stream loading status messages into the response writer
+		// add a sync so the streaming client only runs when the goroutine has exited
+
+		isStreaming, _ := r.Context().Value(proxyCtxKey("streaming")).(bool)
+
+		// PR #417 (no support for anthropic v1/messages yet)
+		isChatCompletions := strings.HasPrefix(r.URL.Path, "/v1/chat/completions")
+		if p.config.SendLoadingState != nil && *p.config.SendLoadingState && isStreaming && isChatCompletions {
+			srw = newStatusResponseWriter(p, w)
+			go srw.statusUpdates(swapCtx)
+		} else {
+			p.proxyLogger.Debugf("<%s> SendLoadingState is nil or false, not streaming loading state", p.ID)
+		}
+
 		beginStartTime := time.Now()
 		if err := p.start(); err != nil {
 			errstr := fmt.Sprintf("unable to start process: %s", err)
-			http.Error(w, errstr, http.StatusBadGateway)
+			cancelLoadCtx()
+			if srw != nil {
+				srw.sendData(fmt.Sprintf("Unable to swap model err: %s\n", errstr))
+				// Wait for statusUpdates goroutine to finish writing its deferred "Done!" messages
+				// before closing the connection. Without this, the connection would close before
+				// the goroutine can write its cleanup messages, causing incomplete SSE output.
+				srw.waitForCompletion(100 * time.Millisecond)
+			} else {
+				http.Error(w, errstr, http.StatusBadGateway)
+			}
 			return
 		}
 		startDuration = time.Since(beginStartTime)
 	}
 
-	if p.reverseProxy != nil {
-		p.reverseProxy.ServeHTTP(w, r)
+	// should trigger srw to stop sending loading events ...
+	cancelLoadCtx()
+
+	// recover from http.ErrAbortHandler panics that can occur when the client
+	// disconnects before the response is sent
+	defer func() {
+		if r := recover(); r != nil {
+			if r == http.ErrAbortHandler {
+				p.proxyLogger.Infof("<%s> recovered from client disconnection during streaming", p.ID)
+			} else {
+				p.proxyLogger.Infof("<%s> recovered from panic: %v", p.ID, r)
+			}
+		}
+	}()
+
+	if srw != nil {
+		// Wait for the goroutine to finish writing its final messages
+		const completionTimeout = 1 * time.Second
+		if !srw.waitForCompletion(completionTimeout) {
+			p.proxyLogger.Warnf("<%s> status updates goroutine did not complete within %v, proceeding with proxy request", p.ID, completionTimeout)
+		}
+		p.reverseProxy.ServeHTTP(srw, r)
 	} else {
-		http.Error(w, fmt.Sprintf("No reverse proxy available for %s", p.ID), http.StatusInternalServerError)
+		p.reverseProxy.ServeHTTP(w, r)
 	}
 
 	totalTime := time.Since(requestBeginTime)
@@ -573,6 +632,7 @@ func (p *Process) cmdStopUpstreamProcess() error {
 		stopCmd := exec.Command(stopArgs[0], stopArgs[1:]...)
 		stopCmd.Stdout = p.processLogger
 		stopCmd.Stderr = p.processLogger
+		setProcAttributes(stopCmd)
 		stopCmd.Env = p.cmd.Env
 
 		if err := stopCmd.Run(); err != nil {
@@ -588,3 +648,232 @@ func (p *Process) cmdStopUpstreamProcess() error {
 
 	return nil
 }
+
+// Logger returns the logger for this process.
+func (p *Process) Logger() *LogMonitor {
+	return p.processLogger
+}
+
+var loadingRemarks = []string{
+	"Still faster than your last standup meeting...",
+	"Reticulating splines...",
+	"Waking up the hamsters...",
+	"Teaching the model manners...",
+	"Convincing the GPU to participate...",
+	"Loading weights (they're heavy)...",
+	"Herding electrons...",
+	"Compiling excuses for the delay...",
+	"Downloading more RAM...",
+	"Asking the model nicely to boot up...",
+	"Bribing CUDA with cookies...",
+	"Still loading (blame VRAM)...",
+	"The model is fashionably late...",
+	"Warming up those tensors...",
+	"Making the neural net do push-ups...",
+	"Your patience is appreciated (really)...",
+	"Almost there (probably)...",
+	"Loading like it's 1999...",
+	"The model forgot where it put its keys...",
+	"Quantum tunneling through layers...",
+	"Negotiating with the PCIe bus...",
+	"Defrosting frozen parameters...",
+	"Teaching attention heads to focus...",
+	"Running the matrix (slowly)...",
+	"Untangling transformer blocks...",
+	"Calibrating the flux capacitor...",
+	"Spinning up the probability wheels...",
+	"Waiting for the GPU to wake from its nap...",
+	"Converting caffeine to compute...",
+	"Allocating virtual patience...",
+	"Performing arcane CUDA rituals...",
+	"The model is stuck in traffic...",
+	"Inflating embeddings...",
+	"Summoning computational demons...",
+	"Pleading with the OOM killer...",
+	"Calculating the meaning of life (still at 42)...",
+	"Training the training wheels...",
+	"Optimizing the optimizer...",
+	"Bootstrapping the bootstrapper...",
+	"Loading loading screen...",
+	"Processing processing logs...",
+	"Buffering buffer overflow jokes...",
+	"The model hit snooze...",
+	"Debugging the debugger...",
+	"Compiling the compiler...",
+	"Parsing the parser (meta)...",
+	"Tokenizing tokens...",
+	"Encoding the encoder...",
+	"Hashing hash browns...",
+	"Forking spoons (not forks)...",
+	"The model is contemplating existence...",
+	"Transcending dimensional barriers...",
+	"Invoking elder tensor gods...",
+	"Unfurling probability clouds...",
+	"Synchronizing parallel universes...",
+	"The GPU is having second thoughts...",
+	"Recalibrating reality matrices...",
+	"Time is an illusion, loading doubly so...",
+	"Convincing bits to flip themselves...",
+	"The model is reading its own documentation...",
+}
+
+type statusResponseWriter struct {
+	hasWritten bool
+	writer     http.ResponseWriter
+	process    *Process
+	wg         sync.WaitGroup // Track goroutine completion
+	start      time.Time
+}
+
+func newStatusResponseWriter(p *Process, w http.ResponseWriter) *statusResponseWriter {
+	s := &statusResponseWriter{
+		writer:  w,
+		process: p,
+		start:   time.Now(),
+	}
+
+	s.Header().Set("Content-Type", "text/event-stream") // SSE
+	s.Header().Set("Cache-Control", "no-cache")         // no-cache
+	s.Header().Set("Connection", "keep-alive")          // keep-alive
+	s.WriteHeader(http.StatusOK)                        // send status code 200
+	s.sendLine("━━━━━")
+	s.sendLine(fmt.Sprintf("llama-swap loading model: %s", p.ID))
+	return s
+}
+
+// statusUpdates sends status updates to the client while the model is loading
+func (s *statusResponseWriter) statusUpdates(ctx context.Context) {
+	s.wg.Add(1)
+	defer s.wg.Done()
+
+	// Recover from panics caused by client disconnection
+	// Note: recover() only works within the same goroutine, so we need it here
+	defer func() {
+		if r := recover(); r != nil {
+			s.process.proxyLogger.Debugf("<%s> statusUpdates recovered from panic (likely client disconnect): %v", s.process.ID, r)
+		}
+	}()
+
+	defer func() {
+		duration := time.Since(s.start)
+		s.sendLine(fmt.Sprintf("\nDone! (%.2fs)", duration.Seconds()))
+		s.sendLine("━━━━━")
+		s.sendLine(" ")
+	}()
+
+	// Create a shuffled copy of loadingRemarks
+	remarks := make([]string, len(loadingRemarks))
+	copy(remarks, loadingRemarks)
+	rand.Shuffle(len(remarks), func(i, j int) {
+		remarks[i], remarks[j] = remarks[j], remarks[i]
+	})
+	ri := 0
+
+	// Pick a random duration to send a remark
+	nextRemarkIn := time.Duration(2+rand.Intn(4)) * time.Second
+	lastRemarkTime := time.Now()
+
+	ticker := time.NewTicker(time.Second)
+	defer ticker.Stop() // Ensure ticker is stopped to prevent resource leak
+	for {
+		select {
+		case <-ctx.Done():
+			return
+		case <-ticker.C:
+			if s.process.CurrentState() == StateReady {
+				return
+			}
+
+			// Check if it's time for a snarky remark
+			if time.Since(lastRemarkTime) >= nextRemarkIn {
+				remark := remarks[ri%len(remarks)]
+				ri++
+				s.sendLine(fmt.Sprintf("\n%s", remark))
+				lastRemarkTime = time.Now()
+				// Pick a new random duration for the next remark
+				nextRemarkIn = time.Duration(5+rand.Intn(5)) * time.Second
+			} else {
+				s.sendData(".")
+			}
+		}
+	}
+}
+
+// waitForCompletion waits for the statusUpdates goroutine to finish
+func (s *statusResponseWriter) waitForCompletion(timeout time.Duration) bool {
+	done := make(chan struct{})
+	go func() {
+		s.wg.Wait()
+		close(done)
+	}()
+
+	select {
+	case <-done:
+		return true
+	case <-time.After(timeout):
+		return false
+	}
+}
+
+func (s *statusResponseWriter) sendLine(line string) {
+	s.sendData(line + "\n")
+}
+
+func (s *statusResponseWriter) sendData(data string) {
+	// Create the proper SSE JSON structure
+	type Delta struct {
+		ReasoningContent string `json:"reasoning_content"`
+	}
+	type Choice struct {
+		Delta Delta `json:"delta"`
+	}
+	type SSEMessage struct {
+		Choices []Choice `json:"choices"`
+	}
+
+	msg := SSEMessage{
+		Choices: []Choice{
+			{
+				Delta: Delta{
+					ReasoningContent: data,
+				},
+			},
+		},
+	}
+
+	jsonData, err := json.Marshal(msg)
+	if err != nil {
+		s.process.proxyLogger.Errorf("<%s> Failed to marshal SSE message: %v", s.process.ID, err)
+		return
+	}
+
+	// Write SSE formatted data, panic if not able to write
+	_, err = fmt.Fprintf(s.writer, "data: %s\n\n", jsonData)
+	if err != nil {
+		panic(fmt.Sprintf("<%s> Failed to write SSE data: %v", s.process.ID, err))
+	}
+	s.Flush()
+}
+
+func (s *statusResponseWriter) Header() http.Header {
+	return s.writer.Header()
+}
+
+func (s *statusResponseWriter) Write(data []byte) (int, error) {
+	return s.writer.Write(data)
+}
+
+func (s *statusResponseWriter) WriteHeader(statusCode int) {
+	if s.hasWritten {
+		return
+	}
+	s.hasWritten = true
+	s.writer.WriteHeader(statusCode)
+	s.Flush()
+}
+
+func (s *statusResponseWriter) Flush() {
+	if flusher, ok := s.writer.(http.Flusher); ok {
+		flusher.Flush()
+	}
+}

proxy/process_test.go

@@ -494,3 +494,74 @@ func TestProcess_EnvironmentSetCorrectly(t *testing.T) {
 	assert.Equal(t, len(process1.cmd.Environ())+2, len(process2.cmd.Environ()), "process2 should have 2 more environment variables than process1")
 
 }
+
+// TestProcess_ReverseProxyPanicIsHandled tests that panics from
+// httputil.ReverseProxy in Process.ProxyRequest(w, r) do not bubble up and are
+// handled appropriately.
+//
+// httputil.ReverseProxy will panic with http.ErrAbortHandler when it has sent headers
+// can't copy the body. This can be caused by a client disconnecting before the full
+// response is sent from some reason.
+//
+// bug: https://github.com/mostlygeek/llama-swap/issues/362
+// see: https://github.com/golang/go/issues/23643 (where panic was added to httputil.ReverseProxy)
+func TestProcess_ReverseProxyPanicIsHandled(t *testing.T) {
+	// Add defer/recover to catch any panics that aren't handled by ProxyRequest
+	// If this recover() is hit, it means ProxyRequest didn't handle the panic properly
+	defer func() {
+		if r := recover(); r != nil {
+			t.Fatalf("ProxyRequest should handle panics from reverseProxy.ServeHTTP, but panic was not caught: %v", r)
+		}
+	}()
+
+	expectedMessage := "panic_test"
+	config := getTestSimpleResponderConfig(expectedMessage)
+
+	process := NewProcess("panic-test", 5, config, debugLogger, debugLogger)
+	defer process.Stop()
+
+	// Start the process
+	err := process.start()
+	assert.Nil(t, err)
+	assert.Equal(t, StateReady, process.CurrentState())
+
+	// Create a custom ResponseWriter that simulates a client disconnect
+	// by panicking when Write is called after headers are sent
+	panicWriter := &panicOnWriteResponseWriter{
+		ResponseRecorder: httptest.NewRecorder(),
+		shouldPanic:      true,
+	}
+
+	// Make a request that will trigger the panic
+	req := httptest.NewRequest("GET", "/slow-respond?echo=test&delay=100ms", nil)
+
+	// This should panic inside reverseProxy.ServeHTTP when the panicWriter.Write() is called.
+	// ProxyRequest should catch and handle this panic gracefully.
+	process.ProxyRequest(panicWriter, req)
+
+	// If we get here, the panic was properly recovered in ProxyRequest
+	// The process should still be in a ready state
+	assert.Equal(t, StateReady, process.CurrentState())
+}
+
+// panicOnWriteResponseWriter is a ResponseWriter that panics on Write
+// to simulate a client disconnect after headers are sent
+// used by: TestProcess_ReverseProxyPanicIsHandled
+type panicOnWriteResponseWriter struct {
+	*httptest.ResponseRecorder
+	shouldPanic   bool
+	headerWritten bool
+}
+
+func (w *panicOnWriteResponseWriter) WriteHeader(statusCode int) {
+	w.headerWritten = true
+	w.ResponseRecorder.WriteHeader(statusCode)
+}
+
+func (w *panicOnWriteResponseWriter) Write(b []byte) (int, error) {
+	if w.shouldPanic && w.headerWritten {
+		// Simulate the panic that httputil.ReverseProxy throws
+		panic(http.ErrAbortHandler)
+	}
+	return w.ResponseRecorder.Write(b)
+}

proxy/process_unix.go

@@ -0,0 +1,12 @@
+//go:build !windows
+
+package proxy
+
+import (
+	"os/exec"
+)
+
+// setProcAttributes sets platform-specific process attributes
+func setProcAttributes(cmd *exec.Cmd) {
+	// No-op on Unix systems
+}

proxy/process_windows.go

@@ -0,0 +1,16 @@
+//go:build windows
+
+package proxy
+
+import (
+	"os/exec"
+	"syscall"
+)
+
+// setProcAttributes sets platform-specific process attributes
+func setProcAttributes(cmd *exec.Cmd) {
+	cmd.SysProcAttr = &syscall.SysProcAttr{
+		HideWindow:    true,
+		CreationFlags: 0x08000000, // CREATE_NO_WINDOW
+	}
+}

proxy/processgroup.go

@@ -46,7 +46,8 @@ func NewProcessGroup(id string, config config.Config, proxyLogger *LogMonitor, u
 	// Create a Process for each member in the group
 	for _, modelID := range groupConfig.Members {
 		modelConfig, modelID, _ := pg.config.FindConfig(modelID)
-		process := NewProcess(modelID, pg.config.HealthCheckTimeout, modelConfig, pg.upstreamLogger, pg.proxyLogger)
+		processLogger := NewLogMonitorWriter(upstreamLogger)
+		process := NewProcess(modelID, pg.config.HealthCheckTimeout, modelConfig, processLogger, pg.proxyLogger)
 		pg.processes[modelID] = process
 	}
 
@@ -88,6 +89,13 @@ func (pg *ProcessGroup) HasMember(modelName string) bool {
 	return slices.Contains(pg.config.Groups[pg.id].Members, modelName)
 }
 
+func (pg *ProcessGroup) GetMember(modelName string) (*Process, bool) {
+	if pg.HasMember(modelName) {
+		return pg.processes[modelName], true
+	}
+	return nil, false
+}
+
 func (pg *ProcessGroup) StopProcess(modelID string, strategy StopStrategy) error {
 	pg.Lock()
 

proxy/proxymanager.go

@@ -25,6 +25,8 @@ const (
 	PROFILE_SPLIT_CHAR = ":"
 )
 
+type proxyCtxKey string
+
 type ProxyManager struct {
 	sync.Mutex
 
@@ -36,26 +38,51 @@ type ProxyManager struct {
 	upstreamLogger *LogMonitor
 	muxLogger      *LogMonitor
 
-	metricsMonitor *MetricsMonitor
+	metricsMonitor *metricsMonitor
 
 	processGroups map[string]*ProcessGroup
 
 	// shutdown signaling
 	shutdownCtx    context.Context
 	shutdownCancel context.CancelFunc
+
+	// version info
+	buildDate string
+	commit    string
+	version   string
 }
 
-func New(config config.Config) *ProxyManager {
+func New(proxyConfig config.Config) *ProxyManager {
 	// set up loggers
-	stdoutLogger := NewLogMonitorWriter(os.Stdout)
-	upstreamLogger := NewLogMonitorWriter(stdoutLogger)
-	proxyLogger := NewLogMonitorWriter(stdoutLogger)
 
-	if config.LogRequests {
+	var muxLogger, upstreamLogger, proxyLogger *LogMonitor
+	switch proxyConfig.LogToStdout {
+	case config.LogToStdoutNone:
+		muxLogger = NewLogMonitorWriter(io.Discard)
+		upstreamLogger = NewLogMonitorWriter(io.Discard)
+		proxyLogger = NewLogMonitorWriter(io.Discard)
+	case config.LogToStdoutBoth:
+		muxLogger = NewLogMonitorWriter(os.Stdout)
+		upstreamLogger = NewLogMonitorWriter(muxLogger)
+		proxyLogger = NewLogMonitorWriter(muxLogger)
+	case config.LogToStdoutUpstream:
+		muxLogger = NewLogMonitorWriter(os.Stdout)
+		upstreamLogger = NewLogMonitorWriter(muxLogger)
+		proxyLogger = NewLogMonitorWriter(io.Discard)
+	default:
+		// same as config.LogToStdoutProxy
+		// helpful because some old tests create a config.Config directly and it
+		// may not have LogToStdout set explicitly
+		muxLogger = NewLogMonitorWriter(os.Stdout)
+		upstreamLogger = NewLogMonitorWriter(io.Discard)
+		proxyLogger = NewLogMonitorWriter(muxLogger)
+	}
+
+	if proxyConfig.LogRequests {
 		proxyLogger.Warn("LogRequests configuration is deprecated. Use logLevel instead.")
 	}
 
-	switch strings.ToLower(strings.TrimSpace(config.LogLevel)) {
+	switch strings.ToLower(strings.TrimSpace(proxyConfig.LogLevel)) {
 	case "debug":
 		proxyLogger.SetLogLevel(LevelDebug)
 		upstreamLogger.SetLogLevel(LevelDebug)
@@ -73,38 +100,73 @@ func New(config config.Config) *ProxyManager {
 		upstreamLogger.SetLogLevel(LevelInfo)
 	}
 
+	// see: https://go.dev/src/time/format.go
+	timeFormats := map[string]string{
+		"ansic":       time.ANSIC,
+		"unixdate":    time.UnixDate,
+		"rubydate":    time.RubyDate,
+		"rfc822":      time.RFC822,
+		"rfc822z":     time.RFC822Z,
+		"rfc850":      time.RFC850,
+		"rfc1123":     time.RFC1123,
+		"rfc1123z":    time.RFC1123Z,
+		"rfc3339":     time.RFC3339,
+		"rfc3339nano": time.RFC3339Nano,
+		"kitchen":     time.Kitchen,
+		"stamp":       time.Stamp,
+		"stampmilli":  time.StampMilli,
+		"stampmicro":  time.StampMicro,
+		"stampnano":   time.StampNano,
+	}
+
+	if timeFormat, ok := timeFormats[strings.ToLower(strings.TrimSpace(proxyConfig.LogTimeFormat))]; ok {
+		proxyLogger.SetLogTimeFormat(timeFormat)
+		upstreamLogger.SetLogTimeFormat(timeFormat)
+	}
+
 	shutdownCtx, shutdownCancel := context.WithCancel(context.Background())
 
+	var maxMetrics int
+	if proxyConfig.MetricsMaxInMemory <= 0 {
+		maxMetrics = 1000 // Default fallback
+	} else {
+		maxMetrics = proxyConfig.MetricsMaxInMemory
+	}
+
 	pm := &ProxyManager{
-		config:    config,
+		config:    proxyConfig,
 		ginEngine: gin.New(),
 
 		proxyLogger:    proxyLogger,
-		muxLogger:      stdoutLogger,
+		muxLogger:      muxLogger,
 		upstreamLogger: upstreamLogger,
 
-		metricsMonitor: NewMetricsMonitor(&config),
+		metricsMonitor: newMetricsMonitor(proxyLogger, maxMetrics),
 
 		processGroups: make(map[string]*ProcessGroup),
 
 		shutdownCtx:    shutdownCtx,
 		shutdownCancel: shutdownCancel,
+
+		buildDate: "unknown",
+		commit:    "abcd1234",
+		version:   "0",
 	}
 
 	// create the process groups
-	for groupID := range config.Groups {
-		processGroup := NewProcessGroup(groupID, config, proxyLogger, upstreamLogger)
+	for groupID := range proxyConfig.Groups {
+		processGroup := NewProcessGroup(groupID, proxyConfig, proxyLogger, upstreamLogger)
 		pm.processGroups[groupID] = processGroup
 	}
 
 	pm.setupGinEngine()
 
 	// run any startup hooks
-	if len(config.Hooks.OnStartup.Preload) > 0 {
+	if len(proxyConfig.Hooks.OnStartup.Preload) > 0 {
 		// do it in the background, don't block startup -- not sure if good idea yet
 		go func() {
 			discardWriter := &DiscardWriter{}
-			for _, realModelName := range config.Hooks.OnStartup.Preload {
+			for _, realModelName := range proxyConfig.Hooks.OnStartup.Preload {
 				proxyLogger.Infof("Preloading model: %s", realModelName)
 				processGroup, _, err := pm.swapProcessGroup(realModelName)
 
@@ -193,30 +255,30 @@ func (pm *ProxyManager) setupGinEngine() {
 		c.Next()
 	})
 
-	mm := MetricsMiddleware(pm)
-
 	// Set up routes using the Gin engine
-	pm.ginEngine.POST("/v1/chat/completions", mm, pm.proxyOAIHandler)
+	pm.ginEngine.POST("/v1/chat/completions", pm.proxyInferenceHandler)
 	// Support legacy /v1/completions api, see issue #12
-	pm.ginEngine.POST("/v1/completions", mm, pm.proxyOAIHandler)
+	pm.ginEngine.POST("/v1/completions", pm.proxyInferenceHandler)
+	// Support anthropic /v1/messages (added https://github.com/ggml-org/llama.cpp/pull/17570)
+	pm.ginEngine.POST("/v1/messages", pm.proxyInferenceHandler)
 
 	// Support embeddings and reranking
-	pm.ginEngine.POST("/v1/embeddings", mm, pm.proxyOAIHandler)
+	pm.ginEngine.POST("/v1/embeddings", pm.proxyInferenceHandler)
 
 	// llama-server's /reranking endpoint + aliases
-	pm.ginEngine.POST("/reranking", mm, pm.proxyOAIHandler)
-	pm.ginEngine.POST("/rerank", mm, pm.proxyOAIHandler)
-	pm.ginEngine.POST("/v1/rerank", mm, pm.proxyOAIHandler)
-	pm.ginEngine.POST("/v1/reranking", mm, pm.proxyOAIHandler)
+	pm.ginEngine.POST("/reranking", pm.proxyInferenceHandler)
+	pm.ginEngine.POST("/rerank", pm.proxyInferenceHandler)
+	pm.ginEngine.POST("/v1/rerank", pm.proxyInferenceHandler)
+	pm.ginEngine.POST("/v1/reranking", pm.proxyInferenceHandler)
 
 	// llama-server's /infill endpoint for code infilling
-	pm.ginEngine.POST("/infill", mm, pm.proxyOAIHandler)
+	pm.ginEngine.POST("/infill", pm.proxyInferenceHandler)
 
 	// llama-server's /completion endpoint
-	pm.ginEngine.POST("/completion", mm, pm.proxyOAIHandler)
+	pm.ginEngine.POST("/completion", pm.proxyInferenceHandler)
 
 	// Support audio/speech endpoint
-	pm.ginEngine.POST("/v1/audio/speech", pm.proxyOAIHandler)
+	pm.ginEngine.POST("/v1/audio/speech", pm.proxyInferenceHandler)
 	pm.ginEngine.POST("/v1/audio/transcriptions", pm.proxyOAIPostFormHandler)
 
 	pm.ginEngine.GET("/v1/models", pm.listModelsHandler)
@@ -224,7 +286,7 @@ func (pm *ProxyManager) setupGinEngine() {
 	// in proxymanager_loghandlers.go
 	pm.ginEngine.GET("/logs", pm.sendLogsHandlers)
 	pm.ginEngine.GET("/logs/stream", pm.streamLogsHandler)
-	pm.ginEngine.GET("/logs/stream/:logMonitorID", pm.streamLogsHandler)
+	pm.ginEngine.GET("/logs/stream/*logMonitorID", pm.streamLogsHandler)
 
 	/**
 	 * User Interface Endpoints
@@ -369,28 +431,40 @@ func (pm *ProxyManager) listModelsHandler(c *gin.Context) {
 			continue
 		}
 
-		record := gin.H{
-			"id":       id,
-			"object":   "model",
-			"created":  createdTime,
-			"owned_by": "llama-swap",
+		newRecord := func(modelId string) gin.H {
+			record := gin.H{
+				"id":       modelId,
+				"object":   "model",
+				"created":  createdTime,
+				"owned_by": "llama-swap",
+			}
+
+			if name := strings.TrimSpace(modelConfig.Name); name != "" {
+				record["name"] = name
+			}
+			if desc := strings.TrimSpace(modelConfig.Description); desc != "" {
+				record["description"] = desc
+			}
+
+			// Add metadata if present
+			if len(modelConfig.Metadata) > 0 {
+				record["meta"] = gin.H{
+					"llamaswap": modelConfig.Metadata,
+				}
+			}
+			return record
 		}
 
-		if name := strings.TrimSpace(modelConfig.Name); name != "" {
-			record["name"] = name
-		}
-		if desc := strings.TrimSpace(modelConfig.Description); desc != "" {
-			record["description"] = desc
-		}
+		data = append(data, newRecord(id))
 
-		// Add metadata if present
-		if len(modelConfig.Metadata) > 0 {
-			record["meta"] = gin.H{
-				"llamaswap": modelConfig.Metadata,
+		// Include aliases
+		if pm.config.IncludeAliasesInList {
+			for _, alias := range modelConfig.Aliases {
+				if alias := strings.TrimSpace(alias); alias != "" {
+					data = append(data, newRecord(alias))
+				}
 			}
 		}
-
-		data = append(data, record)
 	}
 
 	// Sort by the "id" key
@@ -412,61 +486,61 @@ func (pm *ProxyManager) listModelsHandler(c *gin.Context) {
 	})
 }
 
-func (pm *ProxyManager) proxyToUpstream(c *gin.Context) {
-	upstreamPath := c.Param("upstreamPath")
-
-	// split the upstream path by / and search for the model name
-	parts := strings.Split(strings.TrimSpace(upstreamPath), "/")
-	if len(parts) == 0 {
-		pm.sendErrorResponse(c, http.StatusBadRequest, "model id required in path")
-		return
-	}
-
-	modelFound := false
+// findModelInPath searches for a valid model name in a path with slashes.
+// It iteratively builds up path segments until it finds a matching model.
+// Returns: (searchModelName, realModelName, remainingPath, found)
+// Example: "/author/model/endpoint" with model "author/model" -> ("author/model", "author/model", "/endpoint", true)
+func (pm *ProxyManager) findModelInPath(path string) (searchName string, realName string, remainingPath string, found bool) {
+	parts := strings.Split(strings.TrimSpace(path), "/")
 	searchModelName := ""
-	var modelName, remainingPath string
+
 	for i, part := range parts {
-		if parts[i] == "" {
+		if part == "" {
 			continue
 		}
 
 		if searchModelName == "" {
 			searchModelName = part
 		} else {
-			searchModelName = searchModelName + "/" + parts[i]
+			searchModelName = searchModelName + "/" + part
 		}
 
 		if real, ok := pm.config.RealModelName(searchModelName); ok {
-			modelName = real
-			remainingPath = "/" + strings.Join(parts[i+1:], "/")
-			modelFound = true
-
-			// Check if this is exactly a model name with no additional path
-			// and doesn't end with a trailing slash
-			if remainingPath == "/" && !strings.HasSuffix(upstreamPath, "/") {
-				// Build new URL with query parameters preserved
-				newPath := "/upstream/" + searchModelName + "/"
-				if c.Request.URL.RawQuery != "" {
-					newPath += "?" + c.Request.URL.RawQuery
-				}
-
-				// Use 308 for non-GET/HEAD requests to preserve method
-				if c.Request.Method == http.MethodGet || c.Request.Method == http.MethodHead {
-					c.Redirect(http.StatusMovedPermanently, newPath)
-				} else {
-					c.Redirect(http.StatusPermanentRedirect, newPath)
-				}
-				return
-			}
-			break
+			return searchModelName, real, "/" + strings.Join(parts[i+1:], "/"), true
 		}
 	}
 
+	return "", "", "", false
+}
+
+func (pm *ProxyManager) proxyToUpstream(c *gin.Context) {
+	upstreamPath := c.Param("upstreamPath")
+
+	searchModelName, modelName, remainingPath, modelFound := pm.findModelInPath(upstreamPath)
+
 	if !modelFound {
 		pm.sendErrorResponse(c, http.StatusBadRequest, "model id required in path")
 		return
 	}
 
+	// Check if this is exactly a model name with no additional path
+	// and doesn't end with a trailing slash
+	if remainingPath == "/" && !strings.HasSuffix(upstreamPath, "/") {
+		// Build new URL with query parameters preserved
+		newPath := "/upstream/" + searchModelName + "/"
+		if c.Request.URL.RawQuery != "" {
+			newPath += "?" + c.Request.URL.RawQuery
+		}
+
+		// Use 308 for non-GET/HEAD requests to preserve method
+		if c.Request.Method == http.MethodGet || c.Request.Method == http.MethodHead {
+			c.Redirect(http.StatusMovedPermanently, newPath)
+		} else {
+			c.Redirect(http.StatusPermanentRedirect, newPath)
+		}
+		return
+	}
+
 	processGroup, realModelName, err := pm.swapProcessGroup(modelName)
 	if err != nil {
 		pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error swapping process group: %s", err.Error()))
@@ -474,11 +548,26 @@ func (pm *ProxyManager) proxyToUpstream(c *gin.Context) {
 	}
 
 	// rewrite the path
+	originalPath := c.Request.URL.Path
 	c.Request.URL.Path = remainingPath
-	processGroup.ProxyRequest(realModelName, c.Writer, c.Request)
+
+	// attempt to record metrics if it is a POST request
+	if pm.metricsMonitor != nil && c.Request.Method == "POST" {
+		if err := pm.metricsMonitor.wrapHandler(realModelName, c.Writer, c.Request, processGroup.ProxyRequest); err != nil {
+			pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error proxying metrics wrapped request: %s", err.Error()))
+			pm.proxyLogger.Errorf("Error proxying wrapped upstream request for model %s, path=%s", realModelName, originalPath)
+			return
+		}
+	} else {
+		if err := processGroup.ProxyRequest(realModelName, c.Writer, c.Request); err != nil {
+			pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error proxying request: %s", err.Error()))
+			pm.proxyLogger.Errorf("Error proxying upstream request for model %s, path=%s", realModelName, originalPath)
+			return
+		}
+	}
 }
 
-func (pm *ProxyManager) proxyOAIHandler(c *gin.Context) {
+func (pm *ProxyManager) proxyInferenceHandler(c *gin.Context) {
 	bodyBytes, err := io.ReadAll(c.Request.Body)
 	if err != nil {
 		pm.sendErrorResponse(c, http.StatusBadRequest, "could not ready request body")
@@ -535,10 +624,24 @@ func (pm *ProxyManager) proxyOAIHandler(c *gin.Context) {
 	c.Request.Header.Set("content-length", strconv.Itoa(len(bodyBytes)))
 	c.Request.ContentLength = int64(len(bodyBytes))
 
-	if err := processGroup.ProxyRequest(realModelName, c.Writer, c.Request); err != nil {
-		pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error proxying request: %s", err.Error()))
-		pm.proxyLogger.Errorf("Error Proxying Request for processGroup %s and model %s", processGroup.id, realModelName)
-		return
+	// issue #366 extract values that downstream handlers may need
+	isStreaming := gjson.GetBytes(bodyBytes, "stream").Bool()
+	ctx := context.WithValue(c.Request.Context(), proxyCtxKey("streaming"), isStreaming)
+	ctx = context.WithValue(ctx, proxyCtxKey("model"), realModelName)
+	c.Request = c.Request.WithContext(ctx)
+
+	if pm.metricsMonitor != nil && c.Request.Method == "POST" {
+		if err := pm.metricsMonitor.wrapHandler(realModelName, c.Writer, c.Request, processGroup.ProxyRequest); err != nil {
+			pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error proxying metrics wrapped request: %s", err.Error()))
+			pm.proxyLogger.Errorf("Error Proxying Metrics Wrapped Request for processGroup %s and model %s", processGroup.id, realModelName)
+			return
+		}
+	} else {
+		if err := processGroup.ProxyRequest(realModelName, c.Writer, c.Request); err != nil {
+			pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error proxying request: %s", err.Error()))
+			pm.proxyLogger.Errorf("Error Proxying Request for processGroup %s and model %s", processGroup.id, realModelName)
+			return
+		}
 	}
 }
 
@@ -698,3 +801,11 @@ func (pm *ProxyManager) findGroupByModelName(modelName string) *ProcessGroup {
 	}
 	return nil
 }
+
+func (pm *ProxyManager) SetVersion(buildDate string, commit string, version string) {
+	pm.Lock()
+	defer pm.Unlock()
+	pm.buildDate = buildDate
+	pm.commit = commit
+	pm.version = version
+}

proxy/proxymanager_api.go

@@ -28,6 +28,7 @@ func addApiHandlers(pm *ProxyManager) {
 		apiGroup.POST("/models/unload/*model", pm.apiUnloadSingleModelHandler)
 		apiGroup.GET("/events", pm.apiSendEvents)
 		apiGroup.GET("/metrics", pm.apiGetMetrics)
+		apiGroup.GET("/version", pm.apiGetVersion)
 	}
 }
 
@@ -180,7 +181,7 @@ func (pm *ProxyManager) apiSendEvents(c *gin.Context) {
 	sendLogData("proxy", pm.proxyLogger.GetHistory())
 	sendLogData("upstream", pm.upstreamLogger.GetHistory())
 	sendModels()
-	sendMetrics(pm.metricsMonitor.GetMetrics())
+	sendMetrics(pm.metricsMonitor.getMetrics())
 
 	for {
 		select {
@@ -198,7 +199,7 @@ func (pm *ProxyManager) apiSendEvents(c *gin.Context) {
 }
 
 func (pm *ProxyManager) apiGetMetrics(c *gin.Context) {
-	jsonData, err := pm.metricsMonitor.GetMetricsJSON()
+	jsonData, err := pm.metricsMonitor.getMetricsJSON()
 	if err != nil {
 		c.JSON(http.StatusInternalServerError, gin.H{"error": "failed to get metrics"})
 		return
@@ -227,3 +228,11 @@ func (pm *ProxyManager) apiUnloadSingleModelHandler(c *gin.Context) {
 		c.String(http.StatusOK, "OK")
 	}
 }
+
+func (pm *ProxyManager) apiGetVersion(c *gin.Context) {
+	c.JSON(http.StatusOK, map[string]string{
+		"version":    pm.version,
+		"commit":     pm.commit,
+		"build_date": pm.buildDate,
+	})
+}

proxy/proxymanager_loghandlers.go

@@ -31,7 +31,7 @@ func (pm *ProxyManager) streamLogsHandler(c *gin.Context) {
 	// prevent nginx from buffering streamed logs
 	c.Header("X-Accel-Buffering", "no")
 
-	logMonitorId := c.Param("logMonitorID")
+	logMonitorId := strings.TrimPrefix(c.Param("logMonitorID"), "/")
 	logger, err := pm.getLogger(logMonitorId)
 	if err != nil {
 		c.String(http.StatusBadRequest, err.Error())
@@ -83,18 +83,25 @@ func (pm *ProxyManager) streamLogsHandler(c *gin.Context) {
 
 // getLogger searches for the appropriate logger based on the logMonitorId
 func (pm *ProxyManager) getLogger(logMonitorId string) (*LogMonitor, error) {
-	var logger *LogMonitor
-
-	if logMonitorId == "" {
+	switch logMonitorId {
+	case "":
 		// maintain the default
-		logger = pm.muxLogger
-	} else if logMonitorId == "proxy" {
-		logger = pm.proxyLogger
-	} else if logMonitorId == "upstream" {
-		logger = pm.upstreamLogger
-	} else {
-		return nil, fmt.Errorf("invalid logger. Use 'proxy' or 'upstream'")
-	}
+		return pm.muxLogger, nil
+	case "proxy":
+		return pm.proxyLogger, nil
+	case "upstream":
+		return pm.upstreamLogger, nil
+	default:
+		// search for a models specific logger using findModelInPath
+		// to handle model names with slashes (e.g., "author/model")
+		if _, name, _, found := pm.findModelInPath("/" + logMonitorId); found {
+			for _, group := range pm.processGroups {
+				if process, found := group.GetMember(name); found {
+					return process.Logger(), nil
+				}
+			}
+		}
 
-	return logger, nil
+		return nil, fmt.Errorf("invalid logger. Use 'proxy', 'upstream' or a model's ID")
+	}
 }

proxy/proxymanager_test.go

@@ -437,6 +437,70 @@ func TestProxyManager_ListModelsHandler_SortedByID(t *testing.T) {
 	}
 }
 
+func TestProxyManager_ListModelsHandler_IncludeAliasesInList(t *testing.T) {
+	// Configure alias
+	config := config.Config{
+		HealthCheckTimeout:   15,
+		IncludeAliasesInList: true,
+		Models: map[string]config.ModelConfig{
+			"model1": func() config.ModelConfig {
+				mc := getTestSimpleResponderConfig("model1")
+				mc.Name = "Model 1"
+				mc.Aliases = []string{"alias1"}
+				return mc
+			}(),
+		},
+		LogLevel: "error",
+	}
+
+	proxy := New(config)
+
+	// Request models list
+	req := httptest.NewRequest("GET", "/v1/models", nil)
+	w := CreateTestResponseRecorder()
+	proxy.ServeHTTP(w, req)
+
+	assert.Equal(t, http.StatusOK, w.Code)
+
+	var response struct {
+		Data []map[string]interface{} `json:"data"`
+	}
+	if err := json.Unmarshal(w.Body.Bytes(), &response); err != nil {
+		t.Fatalf("Failed to parse JSON response: %v", err)
+	}
+
+	// We expect both base id and alias
+	var model1Data, alias1Data map[string]any
+	for _, model := range response.Data {
+		if model["id"] == "model1" {
+			model1Data = model
+		} else if model["id"] == "alias1" {
+			alias1Data = model
+		}
+	}
+
+	// Verify model1 has name
+	assert.NotNil(t, model1Data)
+	_, exists := model1Data["name"]
+	if !assert.True(t, exists, "model1 should have name key") {
+		t.FailNow()
+	}
+	name1, ok := model1Data["name"].(string)
+	assert.True(t, ok, "name1 should be a string")
+
+	// Verify alias1 has name
+	assert.NotNil(t, alias1Data)
+	_, exists = alias1Data["name"]
+	if !assert.True(t, exists, "alias1 should have name key") {
+		t.FailNow()
+	}
+	name2, ok := alias1Data["name"].(string)
+	assert.True(t, ok, "name2 should be a string")
+
+	// Name keys should match
+	assert.Equal(t, name1, name2)
+}
+
 func TestProxyManager_Shutdown(t *testing.T) {
 	// make broken model configurations
 	model1Config := getTestSimpleResponderConfigPort("model1", 9991)
@@ -911,76 +975,6 @@ func TestProxyManager_FiltersStripParams(t *testing.T) {
 	// t.Logf("%v", response)
 }
 
-func TestProxyManager_MiddlewareWritesMetrics_NonStreaming(t *testing.T) {
-	config := config.AddDefaultGroupToConfig(config.Config{
-		HealthCheckTimeout: 15,
-		Models: map[string]config.ModelConfig{
-			"model1": getTestSimpleResponderConfig("model1"),
-		},
-		LogLevel: "error",
-	})
-
-	proxy := New(config)
-	defer proxy.StopProcesses(StopWaitForInflightRequest)
-
-	// Make a non-streaming request
-	reqBody := `{"model":"model1", "stream": false}`
-	req := httptest.NewRequest("POST", "/v1/chat/completions", bytes.NewBufferString(reqBody))
-	w := CreateTestResponseRecorder()
-
-	proxy.ServeHTTP(w, req)
-	assert.Equal(t, http.StatusOK, w.Code)
-
-	// Check that metrics were recorded
-	metrics := proxy.metricsMonitor.GetMetrics()
-	if !assert.NotEmpty(t, metrics, "metrics should be recorded for non-streaming request") {
-		return
-	}
-
-	// Verify the last metric has the correct model
-	lastMetric := metrics[len(metrics)-1]
-	assert.Equal(t, "model1", lastMetric.Model)
-	assert.Equal(t, 25, lastMetric.InputTokens, "input tokens should be 25")
-	assert.Equal(t, 10, lastMetric.OutputTokens, "output tokens should be 10")
-	assert.Greater(t, lastMetric.TokensPerSecond, 0.0, "tokens per second should be greater than 0")
-	assert.Greater(t, lastMetric.DurationMs, 0, "duration should be greater than 0")
-}
-
-func TestProxyManager_MiddlewareWritesMetrics_Streaming(t *testing.T) {
-	config := config.AddDefaultGroupToConfig(config.Config{
-		HealthCheckTimeout: 15,
-		Models: map[string]config.ModelConfig{
-			"model1": getTestSimpleResponderConfig("model1"),
-		},
-		LogLevel: "error",
-	})
-
-	proxy := New(config)
-	defer proxy.StopProcesses(StopWaitForInflightRequest)
-
-	// Make a streaming request
-	reqBody := `{"model":"model1", "stream": true}`
-	req := httptest.NewRequest("POST", "/v1/chat/completions?stream=true", bytes.NewBufferString(reqBody))
-	w := CreateTestResponseRecorder()
-
-	proxy.ServeHTTP(w, req)
-	assert.Equal(t, http.StatusOK, w.Code)
-
-	// Check that metrics were recorded
-	metrics := proxy.metricsMonitor.GetMetrics()
-	if !assert.NotEmpty(t, metrics, "metrics should be recorded for streaming request") {
-		return
-	}
-
-	// Verify the last metric has the correct model
-	lastMetric := metrics[len(metrics)-1]
-	assert.Equal(t, "model1", lastMetric.Model)
-	assert.Equal(t, 25, lastMetric.InputTokens, "input tokens should be 25")
-	assert.Equal(t, 10, lastMetric.OutputTokens, "output tokens should be 10")
-	assert.Greater(t, lastMetric.TokensPerSecond, 0.0, "tokens per second should be greater than 0")
-	assert.Greater(t, lastMetric.DurationMs, 0, "duration should be greater than 0")
-}
-
 func TestProxyManager_HealthEndpoint(t *testing.T) {
 	config := config.AddDefaultGroupToConfig(config.Config{
 		HealthCheckTimeout: 15,
@@ -1084,7 +1078,8 @@ func TestProxyManager_StreamingEndpointsReturnNoBufferingHeader(t *testing.T) {
 	config := config.AddDefaultGroupToConfig(config.Config{
 		HealthCheckTimeout: 15,
 		Models: map[string]config.ModelConfig{
-			"model1": getTestSimpleResponderConfig("model1"),
+			"model1":       getTestSimpleResponderConfig("model1"),
+			"author/model": getTestSimpleResponderConfig("author/model"),
 		},
 		LogLevel: "error",
 	})
@@ -1097,6 +1092,7 @@ func TestProxyManager_StreamingEndpointsReturnNoBufferingHeader(t *testing.T) {
 		"/logs/stream",
 		"/logs/stream/proxy",
 		"/logs/stream/upstream",
+		"/logs/stream/author/model",
 	}
 
 	for _, endpoint := range endpoints {
@@ -1153,3 +1149,41 @@ func TestProxyManager_ProxiedStreamingEndpointReturnsNoBufferingHeader(t *testin
 	assert.Equal(t, "no", rec.Header().Get("X-Accel-Buffering"))
 	assert.Contains(t, rec.Header().Get("Content-Type"), "text/event-stream")
 }
+
+func TestProxyManager_ApiGetVersion(t *testing.T) {
+	config := config.AddDefaultGroupToConfig(config.Config{
+		HealthCheckTimeout: 15,
+		Models: map[string]config.ModelConfig{
+			"model1": getTestSimpleResponderConfig("model1"),
+		},
+		LogLevel: "error",
+	})
+
+	// Version test map
+	versionTest := map[string]string{
+		"build_date": "1970-01-01T00:00:00Z",
+		"commit":     "cc915ddb6f04a42d9cd1f524e1d46ec6ed069fdc",
+		"version":    "v001",
+	}
+
+	proxy := New(config)
+	proxy.SetVersion(versionTest["build_date"], versionTest["commit"], versionTest["version"])
+	defer proxy.StopProcesses(StopWaitForInflightRequest)
+
+	req := httptest.NewRequest("GET", "/api/version", nil)
+	w := CreateTestResponseRecorder()
+
+	proxy.ServeHTTP(w, req)
+
+	assert.Equal(t, http.StatusOK, w.Code)
+
+	// Ensure json response
+	assert.Equal(t, "application/json; charset=utf-8", w.Header().Get("Content-Type"))
+
+	// Check for attributes
+	response := map[string]string{}
+	assert.NoError(t, json.Unmarshal(w.Body.Bytes(), &response))
+	for key, value := range versionTest {
+		assert.Equal(t, value, response[key], "%s value %s should match response %s", key, value, response[key])
+	}
+}

ui/package-lock.json

@@ -752,9 +752,9 @@
       }
     },
     "node_modules/@eslint-community/eslint-utils": {
-      "version": "4.7.0",
-      "resolved": "https://registry.npmjs.org/@eslint-community/eslint-utils/-/eslint-utils-4.7.0.tgz",
-      "integrity": "sha512-dyybb3AcajC7uha6CvhdVRJqaKyn7w2YKqKyAN37NKYgZT36w+iRb0Dymmc5qEJ549c/S31cMMSFd75bteCpCw==",
+      "version": "4.9.0",
+      "resolved": "https://registry.npmjs.org/@eslint-community/eslint-utils/-/eslint-utils-4.9.0.tgz",
+      "integrity": "sha512-ayVFHdtZ+hsq1t2Dy24wCmGXGe4q9Gu3smhLYALJrr473ZH27MsnSL+LKUlimp4BWJqMDMLmPpx/Q9R3OAlL4g==",
       "dev": true,
       "license": "MIT",
       "dependencies": {
@@ -794,13 +794,13 @@
       }
     },
     "node_modules/@eslint/config-array": {
-      "version": "0.20.0",
-      "resolved": "https://registry.npmjs.org/@eslint/config-array/-/config-array-0.20.0.tgz",
-      "integrity": "sha512-fxlS1kkIjx8+vy2SjuCB94q3htSNrufYTXubwiBFeaQHbH6Ipi43gFJq2zCMt6PHhImH3Xmr0NksKDvchWlpQQ==",
+      "version": "0.21.1",
+      "resolved": "https://registry.npmjs.org/@eslint/config-array/-/config-array-0.21.1.tgz",
+      "integrity": "sha512-aw1gNayWpdI/jSYVgzN5pL0cfzU02GT3NBpeT/DXbx1/1x7ZKxFPd9bwrzygx/qiwIQiJ1sw/zD8qY/kRvlGHA==",
       "dev": true,
       "license": "Apache-2.0",
       "dependencies": {
-        "@eslint/object-schema": "^2.1.6",
+        "@eslint/object-schema": "^2.1.7",
         "debug": "^4.3.1",
         "minimatch": "^3.1.2"
       },
@@ -809,19 +809,22 @@
       }
     },
     "node_modules/@eslint/config-helpers": {
-      "version": "0.2.2",
-      "resolved": "https://registry.npmjs.org/@eslint/config-helpers/-/config-helpers-0.2.2.tgz",
-      "integrity": "sha512-+GPzk8PlG0sPpzdU5ZvIRMPidzAnZDl/s9L+y13iodqvb8leL53bTannOrQ/Im7UkpsmFU5Ily5U60LWixnmLg==",
+      "version": "0.4.2",
+      "resolved": "https://registry.npmjs.org/@eslint/config-helpers/-/config-helpers-0.4.2.tgz",
+      "integrity": "sha512-gBrxN88gOIf3R7ja5K9slwNayVcZgK6SOUORm2uBzTeIEfeVaIhOpCtTox3P6R7o2jLFwLFTLnC7kU/RGcYEgw==",
       "dev": true,
       "license": "Apache-2.0",
+      "dependencies": {
+        "@eslint/core": "^0.17.0"
+      },
       "engines": {
         "node": "^18.18.0 || ^20.9.0 || >=21.1.0"
       }
     },
     "node_modules/@eslint/core": {
-      "version": "0.14.0",
-      "resolved": "https://registry.npmjs.org/@eslint/core/-/core-0.14.0.tgz",
-      "integrity": "sha512-qIbV0/JZr7iSDjqAc60IqbLdsj9GDt16xQtWD+B78d/HAlvysGdZZ6rpJHGAc2T0FQx1X6thsSPdnoiGKdNtdg==",
+      "version": "0.17.0",
+      "resolved": "https://registry.npmjs.org/@eslint/core/-/core-0.17.0.tgz",
+      "integrity": "sha512-yL/sLrpmtDaFEiUj1osRP4TI2MDz1AddJL+jZ7KSqvBuliN4xqYY54IfdN8qD8Toa6g1iloph1fxQNkjOxrrpQ==",
       "dev": true,
       "license": "Apache-2.0",
       "dependencies": {
@@ -869,9 +872,9 @@
       }
     },
     "node_modules/@eslint/js": {
-      "version": "9.28.0",
-      "resolved": "https://registry.npmjs.org/@eslint/js/-/js-9.28.0.tgz",
-      "integrity": "sha512-fnqSjGWd/CoIp4EXIxWVK/sHA6DOHN4+8Ix2cX5ycOY7LG0UY8nHCU5pIp2eaE1Mc7Qd8kHspYNzYXT2ojPLzg==",
+      "version": "9.39.1",
+      "resolved": "https://registry.npmjs.org/@eslint/js/-/js-9.39.1.tgz",
+      "integrity": "sha512-S26Stp4zCy88tH94QbBv3XCuzRQiZ9yXofEILmglYTh/Ug/a9/umqvgFtYBAo3Lp0nsI/5/qH1CCrbdK3AP1Tw==",
       "dev": true,
       "license": "MIT",
       "engines": {
@@ -882,9 +885,9 @@
       }
     },
     "node_modules/@eslint/object-schema": {
-      "version": "2.1.6",
-      "resolved": "https://registry.npmjs.org/@eslint/object-schema/-/object-schema-2.1.6.tgz",
-      "integrity": "sha512-RBMg5FRL0I0gs51M/guSAj5/e14VQ4tpZnQNWwuDT66P14I43ItmPfIZRhO9fUVIPOAQXU47atlywZ/czoqFPA==",
+      "version": "2.1.7",
+      "resolved": "https://registry.npmjs.org/@eslint/object-schema/-/object-schema-2.1.7.tgz",
+      "integrity": "sha512-VtAOaymWVfZcmZbp6E2mympDIHvyjXs/12LqWYjVw6qjrfF+VK+fyG33kChz3nnK+SU5/NeHOqrTEHS8sXO3OA==",
       "dev": true,
       "license": "Apache-2.0",
       "engines": {
@@ -892,13 +895,13 @@
       }
     },
     "node_modules/@eslint/plugin-kit": {
-      "version": "0.3.1",
-      "resolved": "https://registry.npmjs.org/@eslint/plugin-kit/-/plugin-kit-0.3.1.tgz",
-      "integrity": "sha512-0J+zgWxHN+xXONWIyPWKFMgVuJoZuGiIFu8yxk7RJjxkzpGmyja5wRFqZIVtjDVOQpV+Rw0iOAjYPE2eQyjr0w==",
+      "version": "0.4.1",
+      "resolved": "https://registry.npmjs.org/@eslint/plugin-kit/-/plugin-kit-0.4.1.tgz",
+      "integrity": "sha512-43/qtrDUokr7LJqoF2c3+RInu/t4zfrpYdoSDfYyhg52rwLV6TnOvdG4fXm7IkSB3wErkcmJS9iEhjVtOSEjjA==",
       "dev": true,
       "license": "Apache-2.0",
       "dependencies": {
-        "@eslint/core": "^0.14.0",
+        "@eslint/core": "^0.17.0",
         "levn": "^0.4.1"
       },
       "engines": {
@@ -1908,9 +1911,9 @@
       }
     },
     "node_modules/@typescript-eslint/typescript-estree/node_modules/brace-expansion": {
-      "version": "2.0.1",
-      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-2.0.1.tgz",
-      "integrity": "sha512-XnAIvQ8eM+kC6aULx6wuQiwVsnzsi9d3WxzV3FpWTGA19F621kwdbsAcFKXgKUHZWsy+mY6iL1sHTxWEFCytDA==",
+      "version": "2.0.2",
+      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-2.0.2.tgz",
+      "integrity": "sha512-Jt0vHyM+jmUBqojB7E1NIYadt0vI0Qxjxd2TErW94wDz+E2LAm5vKMXXwg6ZZBTHPuUlDgQHKXvjGBdfcF1ZDQ==",
       "dev": true,
       "license": "MIT",
       "dependencies": {
@@ -2010,9 +2013,9 @@
       }
     },
     "node_modules/acorn": {
-      "version": "8.14.1",
-      "resolved": "https://registry.npmjs.org/acorn/-/acorn-8.14.1.tgz",
-      "integrity": "sha512-OvQ/2pUDKmgfCg++xsTX1wGxfTaszcHVcTctW4UJB4hibJx2HXxxO5UmVgyjMa+ZDsiaf5wWLXYpRWMmBI0QHg==",
+      "version": "8.15.0",
+      "resolved": "https://registry.npmjs.org/acorn/-/acorn-8.15.0.tgz",
+      "integrity": "sha512-NZyJarBfL7nWwIq+FDL6Zp/yHEhePMNnnJ0y3qfieCrmNvYct8uvtiV41UvlSe6apAfk0fY1FbWx+NwfmpvtTg==",
       "dev": true,
       "license": "MIT",
       "bin": {
@@ -2080,9 +2083,9 @@
       "license": "MIT"
     },
     "node_modules/brace-expansion": {
-      "version": "1.1.11",
-      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-1.1.11.tgz",
-      "integrity": "sha512-iCuPHDFgrHX7H2vEI/5xpz07zSHB00TpugqhmYtVmMO6518mCuRMoOYFldEBl0g187ufozdaHgWKcYFb61qGiA==",
+      "version": "1.1.12",
+      "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-1.1.12.tgz",
+      "integrity": "sha512-9T9UjW3r0UW5c1Q7GTwllptXwhvYmEzFhzMfZ9H7FQWt+uZePjZPjBP/W1ZEyZ1twGWom5/56TF4lPcqjnDHcg==",
       "dev": true,
       "license": "MIT",
       "dependencies": {
@@ -2380,33 +2383,32 @@
       }
     },
     "node_modules/eslint": {
-      "version": "9.28.0",
-      "resolved": "https://registry.npmjs.org/eslint/-/eslint-9.28.0.tgz",
-      "integrity": "sha512-ocgh41VhRlf9+fVpe7QKzwLj9c92fDiqOj8Y3Sd4/ZmVA4Btx4PlUYPq4pp9JDyupkf1upbEXecxL2mwNV7jPQ==",
+      "version": "9.39.1",
+      "resolved": "https://registry.npmjs.org/eslint/-/eslint-9.39.1.tgz",
+      "integrity": "sha512-BhHmn2yNOFA9H9JmmIVKJmd288g9hrVRDkdoIgRCRuSySRUHH7r/DI6aAXW9T1WwUuY3DFgrcaqB+deURBLR5g==",
       "dev": true,
       "license": "MIT",
       "dependencies": {
-        "@eslint-community/eslint-utils": "^4.2.0",
+        "@eslint-community/eslint-utils": "^4.8.0",
         "@eslint-community/regexpp": "^4.12.1",
-        "@eslint/config-array": "^0.20.0",
-        "@eslint/config-helpers": "^0.2.1",
-        "@eslint/core": "^0.14.0",
+        "@eslint/config-array": "^0.21.1",
+        "@eslint/config-helpers": "^0.4.2",
+        "@eslint/core": "^0.17.0",
         "@eslint/eslintrc": "^3.3.1",
-        "@eslint/js": "9.28.0",
-        "@eslint/plugin-kit": "^0.3.1",
+        "@eslint/js": "9.39.1",
+        "@eslint/plugin-kit": "^0.4.1",
         "@humanfs/node": "^0.16.6",
         "@humanwhocodes/module-importer": "^1.0.1",
         "@humanwhocodes/retry": "^0.4.2",
         "@types/estree": "^1.0.6",
-        "@types/json-schema": "^7.0.15",
         "ajv": "^6.12.4",
         "chalk": "^4.0.0",
         "cross-spawn": "^7.0.6",
         "debug": "^4.3.2",
         "escape-string-regexp": "^4.0.0",
-        "eslint-scope": "^8.3.0",
-        "eslint-visitor-keys": "^4.2.0",
-        "espree": "^10.3.0",
+        "eslint-scope": "^8.4.0",
+        "eslint-visitor-keys": "^4.2.1",
+        "espree": "^10.4.0",
         "esquery": "^1.5.0",
         "esutils": "^2.0.2",
         "fast-deep-equal": "^3.1.3",
@@ -2464,9 +2466,9 @@
       }
     },
     "node_modules/eslint-scope": {
-      "version": "8.3.0",
-      "resolved": "https://registry.npmjs.org/eslint-scope/-/eslint-scope-8.3.0.tgz",
-      "integrity": "sha512-pUNxi75F8MJ/GdeKtVLSbYg4ZI34J6C0C7sbL4YOp2exGwen7ZsuBqKzUhXd0qMQ362yET3z+uPwKeg/0C2XCQ==",
+      "version": "8.4.0",
+      "resolved": "https://registry.npmjs.org/eslint-scope/-/eslint-scope-8.4.0.tgz",
+      "integrity": "sha512-sNXOfKCn74rt8RICKMvJS7XKV/Xk9kA7DyJr8mJik3S7Cwgy3qlkkmyS2uQB3jiJg6VNdZd/pDBJu0nvG2NlTg==",
       "dev": true,
       "license": "BSD-2-Clause",
       "dependencies": {
@@ -2481,9 +2483,9 @@
       }
     },
     "node_modules/eslint-visitor-keys": {
-      "version": "4.2.0",
-      "resolved": "https://registry.npmjs.org/eslint-visitor-keys/-/eslint-visitor-keys-4.2.0.tgz",
-      "integrity": "sha512-UyLnSehNt62FFhSwjZlHmeokpRK59rcz29j+F1/aDgbkbRTk7wIc9XzdoasMUbRNKDM0qQt/+BJ4BrpFeABemw==",
+      "version": "4.2.1",
+      "resolved": "https://registry.npmjs.org/eslint-visitor-keys/-/eslint-visitor-keys-4.2.1.tgz",
+      "integrity": "sha512-Uhdk5sfqcee/9H/rCOJikYz67o0a2Tw2hGRPOG2Y1R2dg7brRe1uG0yaNQDHu+TO/uQPF/5eCapvYSmHUjt7JQ==",
       "dev": true,
       "license": "Apache-2.0",
       "engines": {
@@ -2494,15 +2496,15 @@
       }
     },
     "node_modules/espree": {
-      "version": "10.3.0",
-      "resolved": "https://registry.npmjs.org/espree/-/espree-10.3.0.tgz",
-      "integrity": "sha512-0QYC8b24HWY8zjRnDTL6RiHfDbAWn63qb4LMj1Z4b076A4une81+z03Kg7l7mn/48PUTqoLptSXez8oknU8Clg==",
+      "version": "10.4.0",
+      "resolved": "https://registry.npmjs.org/espree/-/espree-10.4.0.tgz",
+      "integrity": "sha512-j6PAQ2uUr79PZhBjP5C5fhl8e39FmRnOjsD5lGnWrFU8i2G776tBK7+nP8KuQUTTyAZUwfQqXAgrVH5MbH9CYQ==",
       "dev": true,
       "license": "BSD-2-Clause",
       "dependencies": {
-        "acorn": "^8.14.0",
+        "acorn": "^8.15.0",
         "acorn-jsx": "^5.3.2",
-        "eslint-visitor-keys": "^4.2.0"
+        "eslint-visitor-keys": "^4.2.1"
       },
       "engines": {
         "node": "^18.18.0 || ^20.9.0 || >=21.1.0"
@@ -2852,9 +2854,9 @@
       "license": "MIT"
     },
     "node_modules/js-yaml": {
-      "version": "4.1.0",
-      "resolved": "https://registry.npmjs.org/js-yaml/-/js-yaml-4.1.0.tgz",
-      "integrity": "sha512-wpxZs9NoxZaJESJGIZTyDEaYpl0FKSA+FB9aJiyemKhMwkxQg63h4T1KJgUGHpTqPDNRcmmYLugrRjJlBtWvRA==",
+      "version": "4.1.1",
+      "resolved": "https://registry.npmjs.org/js-yaml/-/js-yaml-4.1.1.tgz",
+      "integrity": "sha512-qQKT4zQxXl8lLwBtHMWwaTcGfFOZviOJet3Oy/xmGk2gZH677CJM9EvtfdSkgWcATZhj/55JZ0rmy3myCT5lsA==",
       "dev": true,
       "license": "MIT",
       "dependencies": {
@@ -3975,9 +3977,9 @@
       }
     },
     "node_modules/vite": {
-      "version": "6.3.5",
-      "resolved": "https://registry.npmjs.org/vite/-/vite-6.3.5.tgz",
-      "integrity": "sha512-cZn6NDFE7wdTpINgs++ZJ4N49W2vRp8LCKrn3Ob1kYNtOo21vfDoaV5GzBfLU4MovSAB8uNRm4jgzVQZ+mBzPQ==",
+      "version": "6.4.1",
+      "resolved": "https://registry.npmjs.org/vite/-/vite-6.4.1.tgz",
+      "integrity": "sha512-+Oxm7q9hDoLMyJOYfUYBuHQo+dkAloi33apOPP56pzj+vsdJDzr+j1NISE5pyaAuKL4A3UD34qd0lx5+kfKp2g==",
       "dev": true,
       "license": "MIT",
       "dependencies": {

ui/src/components/ConnectionStatus.tsx

@@ -2,7 +2,7 @@ import { useAPI } from "../contexts/APIProvider";
 import { useMemo } from "react";
 
 const ConnectionStatusIcon = () => {
-  const { connectionStatus } = useAPI();
+  const { connectionStatus, versionInfo } = useAPI();
 
   const eventStatusColor = useMemo(() => {
     switch (connectionStatus) {
@@ -17,7 +17,7 @@ const ConnectionStatusIcon = () => {
   }, [connectionStatus]);
 
   return (
-    <div className="flex items-center" title={`event stream: ${connectionStatus}`}>
+    <div className="flex items-center" title={`Event Stream: ${connectionStatus ?? 'unknown'}\nAPI Version: ${versionInfo?.version ?? 'unknown'}\nCommit Hash: ${versionInfo?.commit?.substring(0,7) ?? 'unknown'}\nBuild Date: ${versionInfo?.build_date ?? 'unknown'}`}>
       <span className={`inline-block w-3 h-3 rounded-full ${eventStatusColor} mr-2`}></span>
     </div>
   );

ui/src/contexts/APIProvider.tsx

@@ -23,6 +23,7 @@ interface APIProviderType {
   upstreamLogs: string;
   metrics: Metrics[];
   connectionStatus: ConnectionState;
+  versionInfo: VersionInfo;
 }
 
 interface Metrics {
@@ -41,11 +42,18 @@ interface LogData {
   source: "upstream" | "proxy";
   data: string;
 }
+
 interface APIEventEnvelope {
   type: "modelStatus" | "logData" | "metrics";
   data: string;
 }
 
+interface VersionInfo {
+  build_date: string;
+  commit: string;
+  version: string;
+}
+
 const APIContext = createContext<APIProviderType | undefined>(undefined);
 type APIProviderProps = {
   children: ReactNode;
@@ -59,6 +67,11 @@ export function APIProvider({ children, autoStartAPIEvents = true }: APIProvider
   const [upstreamLogs, setUpstreamLogs] = useState("");
   const [metrics, setMetrics] = useState<Metrics[]>([]);
   const [connectionStatus, setConnectionState] = useState<ConnectionState>("disconnected");
+  const [versionInfo, setVersionInfo] = useState<VersionInfo>({
+    build_date: "unknown",
+    commit: "unknown",
+    version: "unknown"
+  });
   //const apiEventSource = useRef<EventSource | null>(null);
 
   const [models, setModels] = useState<Model[]>([]);
@@ -152,6 +165,26 @@ export function APIProvider({ children, autoStartAPIEvents = true }: APIProvider
     connect();
   }, []);
 
+  useEffect(() => {
+    // fetch version 
+    const fetchVersion = async () => {
+      try {
+        const response = await fetch("/api/version");
+        if (!response.ok) {
+          throw new Error(`HTTP error! status: ${response.status}`);
+        }
+        const data: VersionInfo = await response.json();
+        setVersionInfo(data);
+      } catch (error) {
+        console.error(error);
+      }
+    };
+
+    if (connectionStatus === 'connected') {
+      fetchVersion();
+    }
+  }, [connectionStatus]);
+
   useEffect(() => {
     if (autoStartAPIEvents) {
       enableAPIEvents(true);
@@ -230,8 +263,9 @@ export function APIProvider({ children, autoStartAPIEvents = true }: APIProvider
       upstreamLogs,
       metrics,
       connectionStatus,
+      versionInfo,
     }),
-    [models, listModels, unloadAllModels, loadModel, enableAPIEvents, proxyLogs, upstreamLogs, metrics]
+    [models, listModels, unloadAllModels, unloadSingleModel, loadModel, enableAPIEvents, proxyLogs, upstreamLogs, metrics, connectionStatus, versionInfo]
   );
 
   return <APIContext.Provider value={value}>{children}</APIContext.Provider>;

ui/src/pages/Models.tsx

@@ -191,42 +191,300 @@ function ModelsPanel() {
   );
 }
 
+interface HistogramData {
+  bins: number[];
+  min: number;
+  max: number;
+  binSize: number;
+  p99: number;
+  p95: number;
+  p50: number;
+}
+
+function TokenHistogram({ data }: { data: HistogramData }) {
+  const { bins, min, max, p50, p95, p99 } = data;
+  const maxCount = Math.max(...bins);
+
+  const height = 120;
+  const padding = { top: 10, right: 15, bottom: 25, left: 45 };
+
+  // Use viewBox for responsive sizing
+  const viewBoxWidth = 600;
+  const chartWidth = viewBoxWidth - padding.left - padding.right;
+  const chartHeight = height - padding.top - padding.bottom;
+
+  const barWidth = chartWidth / bins.length;
+  const range = max - min;
+
+  // Calculate x position for a given value
+  const getXPosition = (value: number) => {
+    return padding.left + ((value - min) / range) * chartWidth;
+  };
+
+  return (
+    <div className="mt-2 w-full">
+      <svg
+        viewBox={`0 0 ${viewBoxWidth} ${height}`}
+        className="w-full h-auto"
+        preserveAspectRatio="xMidYMid meet"
+      >
+        {/* Y-axis */}
+        <line
+          x1={padding.left}
+          y1={padding.top}
+          x2={padding.left}
+          y2={height - padding.bottom}
+          stroke="currentColor"
+          strokeWidth="1"
+          opacity="0.3"
+        />
+
+        {/* X-axis */}
+        <line
+          x1={padding.left}
+          y1={height - padding.bottom}
+          x2={viewBoxWidth - padding.right}
+          y2={height - padding.bottom}
+          stroke="currentColor"
+          strokeWidth="1"
+          opacity="0.3"
+        />
+
+        {/* Histogram bars */}
+        {bins.map((count, i) => {
+          const barHeight = maxCount > 0 ? (count / maxCount) * chartHeight : 0;
+          const x = padding.left + i * barWidth;
+          const y = height - padding.bottom - barHeight;
+          const binStart = min + i * data.binSize;
+          const binEnd = binStart + data.binSize;
+
+          return (
+            <g key={i}>
+              <rect
+                x={x}
+                y={y}
+                width={Math.max(barWidth - 1, 1)}
+                height={barHeight}
+                fill="currentColor"
+                opacity="0.6"
+                className="text-blue-500 dark:text-blue-400 hover:opacity-90 transition-opacity cursor-pointer"
+              />
+              <title>{`${binStart.toFixed(1)} - ${binEnd.toFixed(1)} tokens/sec\nCount: ${count}`}</title>
+            </g>
+          );
+        })}
+
+        {/* Percentile lines */}
+        <line
+          x1={getXPosition(p50)}
+          y1={padding.top}
+          x2={getXPosition(p50)}
+          y2={height - padding.bottom}
+          stroke="currentColor"
+          strokeWidth="2"
+          strokeDasharray="4 2"
+          opacity="0.7"
+          className="text-gray-600 dark:text-gray-400"
+        />
+
+        <line
+          x1={getXPosition(p95)}
+          y1={padding.top}
+          x2={getXPosition(p95)}
+          y2={height - padding.bottom}
+          stroke="currentColor"
+          strokeWidth="2"
+          strokeDasharray="4 2"
+          opacity="0.7"
+          className="text-orange-500 dark:text-orange-400"
+        />
+
+        <line
+          x1={getXPosition(p99)}
+          y1={padding.top}
+          x2={getXPosition(p99)}
+          y2={height - padding.bottom}
+          stroke="currentColor"
+          strokeWidth="2"
+          strokeDasharray="4 2"
+          opacity="0.7"
+          className="text-green-500 dark:text-green-400"
+        />
+
+        {/* X-axis labels */}
+        <text
+          x={padding.left}
+          y={height - 5}
+          fontSize="10"
+          fill="currentColor"
+          opacity="0.6"
+          textAnchor="start"
+        >
+          {min.toFixed(1)}
+        </text>
+
+        <text
+          x={viewBoxWidth - padding.right}
+          y={height - 5}
+          fontSize="10"
+          fill="currentColor"
+          opacity="0.6"
+          textAnchor="end"
+        >
+          {max.toFixed(1)}
+        </text>
+
+        {/* X-axis label */}
+        <text
+          x={padding.left + chartWidth / 2}
+          y={height - 2}
+          fontSize="10"
+          fill="currentColor"
+          opacity="0.6"
+          textAnchor="middle"
+        >
+          Tokens/Second Distribution
+        </text>
+      </svg>
+    </div>
+  );
+}
+
 function StatsPanel() {
   const { metrics } = useAPI();
 
-  const [totalRequests, totalInputTokens, totalOutputTokens, avgTokensPerSecond] = useMemo(() => {
+  const [totalRequests, totalInputTokens, totalOutputTokens, tokenStats, histogramData] = useMemo(() => {
     const totalRequests = metrics.length;
     if (totalRequests === 0) {
-      return [0, 0, 0];
+      return [0, 0, 0, { p99: 0, p95: 0, p50: 0 }, null];
     }
     const totalInputTokens = metrics.reduce((sum, m) => sum + m.input_tokens, 0);
     const totalOutputTokens = metrics.reduce((sum, m) => sum + m.output_tokens, 0);
-    const avgTokensPerSecond = (metrics.reduce((sum, m) => sum + m.tokens_per_second, 0) / totalRequests).toFixed(2);
-    return [totalRequests, totalInputTokens, totalOutputTokens, avgTokensPerSecond];
+
+    // Calculate token statistics using output_tokens and duration_ms
+    // Filter out metrics with invalid duration or output tokens
+    const validMetrics = metrics.filter((m) => m.duration_ms > 0 && m.output_tokens > 0);
+    if (validMetrics.length === 0) {
+      return [totalRequests, totalInputTokens, totalOutputTokens, { p99: 0, p95: 0, p50: 0 }, null];
+    }
+
+    // Calculate tokens/second for each valid metric
+    const tokensPerSecond = validMetrics.map((m) => m.output_tokens / (m.duration_ms / 1000));
+
+    // Sort for percentile calculation
+    const sortedTokensPerSecond = [...tokensPerSecond].sort((a, b) => a - b);
+
+    // Calculate percentiles - showing speed thresholds where X% of requests are SLOWER (below)
+    // P99: 99% of requests are slower than this speed (99th percentile - fast requests)
+    // P95: 95% of requests are slower than this speed (95th percentile)
+    // P50: 50% of requests are slower than this speed (median)
+    const p99 = sortedTokensPerSecond[Math.floor(sortedTokensPerSecond.length * 0.99)];
+    const p95 = sortedTokensPerSecond[Math.floor(sortedTokensPerSecond.length * 0.95)];
+    const p50 = sortedTokensPerSecond[Math.floor(sortedTokensPerSecond.length * 0.5)];
+
+    // Create histogram data
+    const min = Math.min(...tokensPerSecond);
+    const max = Math.max(...tokensPerSecond);
+    const binCount = Math.min(30, Math.max(10, Math.floor(tokensPerSecond.length / 5))); // Adaptive bin count
+    const binSize = (max - min) / binCount;
+
+    const bins = Array(binCount).fill(0);
+    tokensPerSecond.forEach((value) => {
+      const binIndex = Math.min(Math.floor((value - min) / binSize), binCount - 1);
+      bins[binIndex]++;
+    });
+
+    const histogramData = {
+      bins,
+      min,
+      max,
+      binSize,
+      p99,
+      p95,
+      p50,
+    };
+
+    return [
+      totalRequests,
+      totalInputTokens,
+      totalOutputTokens,
+      {
+        p99: p99.toFixed(2),
+        p95: p95.toFixed(2),
+        p50: p50.toFixed(2),
+      },
+      histogramData,
+    ];
   }, [metrics]);
 
+  const nf = new Intl.NumberFormat();
+
   return (
     <div className="card">
-      <div className="rounded-lg overflow-hidden border border-gray-200 dark:border-white/10">
-        <table className="w-full">
-          <thead>
-            <tr className="border-b border-gray-200 dark:border-white/10 text-right">
-              <th>Requests</th>
-              <th className="border-l border-gray-200 dark:border-white/10">Processed</th>
-              <th className="border-l border-gray-200 dark:border-white/10">Generated</th>
-              <th className="border-l border-gray-200 dark:border-white/10">Tokens/Sec</th>
+      <div className="rounded-lg overflow-hidden border border-card-border-inner">
+        <table className="min-w-full divide-y divide-card-border-inner">
+          <thead className="bg-secondary">
+            <tr>
+              <th className="px-4 py-3 text-left text-xs font-semibold uppercase tracking-wider text-txtmain">
+                Requests
+              </th>
+              <th className="px-4 py-3 text-left text-xs font-semibold uppercase tracking-wider text-txtmain border-l border-card-border-inner">
+                Processed
+              </th>
+              <th className="px-4 py-3 text-left text-xs font-semibold uppercase tracking-wider text-txtmain border-l border-card-border-inner">
+                Generated
+              </th>
+              <th className="px-4 py-3 text-left text-xs font-semibold uppercase tracking-wider text-txtmain border-l border-card-border-inner">
+                Token Stats (tokens/sec)
+              </th>
             </tr>
           </thead>
-          <tbody>
-            <tr className="text-right">
-              <td className="border-r border-gray-200 dark:border-white/10">{totalRequests}</td>
-              <td className="border-r border-gray-200 dark:border-white/10">
-                {new Intl.NumberFormat().format(totalInputTokens)}
+
+          <tbody className="bg-surface divide-y divide-card-border-inner">
+            <tr className="hover:bg-secondary">
+              <td className="px-4 py-4 text-sm font-semibold text-gray-900 dark:text-white">{totalRequests}</td>
+
+              <td className="px-4 py-4 text-sm text-gray-700 dark:text-gray-300 border-l border-gray-200 dark:border-white/10">
+                <div className="flex items-center gap-2">
+                  <span className="text-sm font-medium">{nf.format(totalInputTokens)}</span>
+                  <span className="text-xs text-gray-500 dark:text-gray-400">tokens</span>
+                </div>
               </td>
-              <td className="border-r border-gray-200 dark:border-white/10">
-                {new Intl.NumberFormat().format(totalOutputTokens)}
+
+              <td className="px-4 py-4 text-sm text-gray-700 dark:text-gray-300 border-l border-gray-200 dark:border-white/10">
+                <div className="flex items-center gap-2">
+                  <span className="text-sm font-medium">{nf.format(totalOutputTokens)}</span>
+                  <span className="text-xs text-gray-500 dark:text-gray-400">tokens</span>
+                </div>
+              </td>
+
+              <td className="px-4 py-4 border-l border-gray-200 dark:border-white/10">
+                <div className="space-y-3">
+                  <div className="grid grid-cols-3 gap-2 items-center">
+                    <div className="text-center">
+                      <div className="text-xs text-gray-500 dark:text-gray-400">P50</div>
+                      <div className="mt-1 inline-block rounded-full bg-gray-100 dark:bg-white/5 px-3 py-1 text-sm font-semibold text-gray-800 dark:text-white">
+                        {tokenStats.p50}
+                      </div>
+                    </div>
+
+                    <div className="text-center">
+                      <div className="text-xs text-gray-500 dark:text-gray-400">P95</div>
+                      <div className="mt-1 inline-block rounded-full bg-gray-100 dark:bg-white/5 px-3 py-1 text-sm font-semibold text-gray-800 dark:text-white">
+                        {tokenStats.p95}
+                      </div>
+                    </div>
+
+                    <div className="text-center">
+                      <div className="text-xs text-gray-500 dark:text-gray-400">P99</div>
+                      <div className="mt-1 inline-block rounded-full bg-gray-100 dark:bg-white/5 px-3 py-1 text-sm font-semibold text-gray-800 dark:text-white">
+                        {tokenStats.p99}
+                      </div>
+                    </div>
+                  </div>
+                  {histogramData && <TokenHistogram data={histogramData} />}
+                </div>
               </td>
-              <td>{avgTokensPerSecond}</td>
             </tr>
           </tbody>
         </table>