API overview

modelux exposes two HTTP APIs:

• Proxy API for running inference. Two surfaces, both authenticated with a project API key:
  • OpenAI-shape at https://api.modelux.ai/openai/v1/* — drop-in replacement for the OpenAI base URL. Routes to OpenAI, Anthropic, Google, Bedrock, Cohere, etc.
  • Anthropic-shape at https://api.modelux.ai/anthropic/v1/* — drop-in replacement for the Anthropic base URL. Same cross-provider routing.
• Management API at https://api.modelux.ai/manage/v1/* — REST API for managing projects, routing configs, providers, budgets, analytics, etc. Authenticated with a management API key.

Both proxy surfaces accept the same model identifiers (raw model names or @config slugs), the same X-Modelux-* headers, and the same routing/budget/observability pipeline. Pick whichever wire format your existing SDK already speaks.

For a per-feature breakdown of what each surface supports and which features round-trip cleanly across each upstream provider, see the capability matrix.

OpenAPI specs

Two specs, served from the docs site so you can codegen clients or import them into Postman / Insomnia:

The proxy spec mirrors the upstream OpenAI / Anthropic shapes for request/response bodies (referencing the official specs for the field catalog) and documents in detail what’s modelux-specific: the X-Modelux-* request headers, the mlx:* request extensions, the response headers (X-Modelux-Request-Id, X-Modelux-Cost-Usd, etc.), and the proxy-only endpoints (/dry-run, /messages/batches, /files, /responses).

Authentication

All API calls require a bearer token:

Authorization: Bearer mlx_sk_...        // proxy / project key
Authorization: Bearer mlx_mk_...        // management key

API key prefixes

Base URLs

Error format

Errors follow the OpenAI error shape for familiarity:

{
  "error": {
    "type": "invalid_request_error",
    "message": "Missing required field: messages",
    "code": "missing_field"
  }
}

Management API errors use the same shape with additional fields where useful:

{
  "error": {
    "type": "not_found",
    "message": "routing config @missing not found",
    "code": "routing_config_not_found",
    "resource": "routing_config",
    "id": "@missing"
  }
}

Status codes

Rate limits

Proxy API limits are per-API-key. Default: 600 requests/minute. Upgrade tiers increase this. Custom limits can be set per key.

Management API: 60 req/min per management key.

Idempotency

Mutating management API endpoints accept an Idempotency-Key header. Same key + same body returns the original response within 24 hours.

Pagination

Management API list endpoints return:

{
  "data": [...],
  "next_cursor": "opaque_cursor_or_null",
  "has_more": true
}

Pass ?cursor=... to fetch the next page.