API Reference

Use our unified REST APIs for models, embeddings, and caching.

Overview

Syndicate Gate provides HTTP endpoints compatible with the OpenAI API. You can update your base URL in the official Python, Node, or Go OpenAI SDKs and enforce requests before provider execution.

Authentication

All endpoints require authentication via a Bearer token issued by your Syndicate Gate control plane:

Authorization: Bearer sk-sg-<your-virtual-api-key>

Text Generation

POST/v1/chat/completions

Create a chat completion. Works identically to the OpenAI Chat Completions API.

{
  "model": "gpt-4o-mini",
  "messages": [
    {"role": "user", "content": "Hello!"}
  ],
  "stream": false
}

POST/v1/messages

Anthropic Messages API compatibility endpoint for SDKs specifically relying on the Anthropic standard.

Error Handling

Errors return standard HTTP status codes and a JSON envelope containing details about the failure:

{
  "error": {
    "message": "Error description",
    "type": "invalid_request_error",
    "param": "model",
    "code": "model_not_found"
  }
}

invalid_request_error: Malformed request or missing parameters.
authentication_error: Invalid or missing API key (or budget exceeded).
rate_limit_error: Organization or Key hit active limits (429).

Request investigation

For request-level investigations in hosted environments, the admin API exposes a request detail endpoint keyed by request ID. This is the same evidence model surfaced in the hosted observability dashboard.

GET /api/v1/admin/requests/:request_id

For the full request-trace model, including hosted and enterprise dashboard surfaces, see Observability & Request Trace.