Ollama Cloud returns HTTP 503 when the model is temporarily overloaded,
429 on rate limit, and 502 on upstream failures. These are transient
conditions that resolve on retry. Previously they bubbled up as hard
errors, forcing callers to implement their own retry logic.
The retry is implemented at the HTTP transport level in doChatRequest,
so both Complete and Stream benefit transparently. Strategy: up to 3
retries with exponential backoff (1s, 2s, 4s), Retry-After header
respected for 429, context cancellation checked between retries.
Non-transient errors (400, 401, 403, 404, 500) are never retried.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Reads Ollama's NDJSON stream (one JSON object per line) and emits
provider.StreamEvent values for text, thinking, tool-call start/delta/end,
and a final Done event carrying assembled Response and Usage. Uses
bufio.Scanner with a 4 MiB max-line buffer so multi-KB tool-call deltas
parse cleanly, and accepts tool-call arguments delivered either as
escaped string fragments (delta-style) or a complete JSON object
(one-shot).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Non-streaming /api/chat support including:
- Vision via images: []base64
- Tool calls on assistant + tool-role response messages
- think field accepting string reasoning levels (or "true"/"false")
- Authorization header when apiKey is non-empty (cloud mode)
Tool-call arguments are passed as JSON objects to the wire and surfaced
as JSON-string Arguments on provider.ToolCall. Tool calls are assigned
synthetic IDs (tc_<index>) when Ollama omits one, so the round-trip
back as an assistant tool_calls + tool-role message remains correlated.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
steve