Chat Completions

POST /v1/chat/completions

OpenAI-compatible request and response shape. Any field the OpenAI API accepts is passed through.

Request

{
  "model": "gpt-4o",
  "messages": [
    { "role": "user", "content": "Summarize for me@acme.com" }
  ],
  "stream": true
}

Request headers

Header	Required	Description
Authorization	Yes	Bearer <api-key>
x-op-team	If key isn't team-scoped	Attributes cost to a team's budget
x-op-policy	No	Comma-separated policy names to apply — see Policies & Guardrails

Response

Standard OpenAI-shaped completion body, plus the x-openproxyai-* response headers documented in the API Reference Overview — cost, latency, provider, cache status, and any policy action taken.

Streaming

Set "stream": true for a server-sent-events stream of ChatCompletionChunk objects, identical in shape to OpenAI's own streaming format.

Errors

See Errors & Status Codes — 402 (budget), 429 (rate limit), 446 (policy block) are specific to OpenProxyAI; everything else follows standard HTTP semantics.