API
API
Mantler exposes deployed stacks as an OpenAI-compatible /v1 API.
The Mantler API is an OpenAI-compatible gateway that routes requests to deployed mantles on your machines.
Base URL: https://api.mantler.ai/v1
Compatibility
The API follows the OpenAI chat completions spec. Any client that works with OpenAI works with Mantler — no changes needed beyond the base URL and API key.
Cursor: Settings → OpenAI → set base URL to https://api.mantler.ai/v1
Continue (VS Code): In config.json, set apiBase: "https://api.mantler.ai/v1"
Supported endpoints
| Endpoint | Description |
|---|---|
GET /health | Gateway health check |
GET /v1/models | List deployed models available to your key |
POST /v1/chat/completions | Generate a chat completion |
POST /v1/completions (legacy completions format) returns an unsupported endpoint error.
In this section
- Authentication — API key format, scopes, org binding
- Endpoints — full request/response spec for each endpoint
- Usage & limits — usage logging and rate limits