API reference · Laras Router Docs

The public API follows the OpenAI Chat Completions format, so the official OpenAI SDKs and most OpenAI-compatible clients work unchanged. It also serves the Anthropic Messages format at POST /v1/messages, for Claude Code and the Anthropic SDKs.

Base URL

Text

https://api.larasrouter.com/v1

All requests use HTTPS. Request and response bodies are JSON (application/json); streamed responses use Server-Sent Events (text/event-stream). Large prompts are supported up to the request size limit; a larger body is rejected with 413 request_too_large.

Authentication

Send your API key in either header:

Headers

Authorization: Bearer lr_live_…
x-api-key: lr_live_…

See Authentication for key management.

Endpoints

GET/v1/modelsList models GET/v1/models/{id}Retrieve a model POST/v1/chat/completionsCreate chat completion POST/v1/messagesCreate message

Generated from the OpenAPI contract, version 0.1.0.

Errors

Every error uses the OpenAI error envelope, except those from POST /v1/messages, which use the Anthropic envelope (see Messages). Branch on error.code, never on the message text: the message is in English by default and in Bahasa Indonesia when the request carries Accept-Language: id.

Error response

{
  "error": {
    "message": "Insufficient balance. Top up to continue.",
    "type": "insufficient_quota",
    "code": "insufficient_credit",
    "param": null
  }
}

errorobjectrequired
The error object.
- messagestringrequired
  Human-readable message. English by default; Bahasa Indonesia with Accept-Language: id. Do not branch on it.
- typestringrequired
  Error category, e.g. invalid_request_error or insufficient_quota.
- codestringrequired
  Stable machine-readable code. Branch on this.
- paramstring | null
  The request parameter the error refers to, or null.

Besides the errors listed on each endpoint, any /v1 path can return these gateway errors:

Status	Code	Type	Meaning
404	`not_found`	`invalid_request_error`	Unknown endpoint. Check the path; the base URL ends in `/v1`.
405	`method_not_allowed`	`invalid_request_error`	The HTTP method is not allowed for this endpoint.
413	`request_too_large`	`invalid_request_error`	The request body is larger than the request size limit. Shorten the prompt or history.
429	`ip_rate_limited`	`rate_limit_exceeded`	Too many requests from one IP address. Wait for `Retry-After`.
500	`internal_error`	`api_error`	Unexpected server error. Retry with backoff.
503	`service_unavailable`	`api_error`	The service is temporarily unavailable. Retry with backoff.

See Errors & retries for what to do about each code.

Response headers

Header	Description
`X-Laras-Request-ID`	Billing identity of the request. Present on responses to requests that reached billing; quote it when you contact support about a charge.
`Retry-After`	On `429` responses: seconds to wait before retrying.

Not available yet

The API serves POST /v1/chat/completions, POST /v1/messages, and GET /v1/models. Other OpenAI endpoints — such as embeddings, images, audio, and the Responses API — are not available. The Responses API (used by Codex CLI) is in development.