Skip to main content

​
API overview

The platform exposes REST endpoints for inference, chat completions, the Agent, fine-tuning, and usage. Authentication is via API key (Bearer token).

​
Base URL

Use this base for inference and agent: 
https://api.llmtune.io
The dashboard and account management are at: 
https://llmtune.io

​
Authentication

All API requests require an API key in the header: 
Authorization: Bearer sk_live_YOUR_KEY
See Authentication for details.

​
Response format

  • Success — JSON body with the requested data (e.g. completion text, usage, job status).
  • Error — JSON with error and optional message. See Errors & status codes.

​
Rate limits

Endpoints are rate-limited per API key. When exceeded, the API returns 429 Too Many Requests; the body may include retryAfter (seconds). Use exponential backoff and avoid tight retry loops.

​
Idempotency

For non-idempotent operations (e.g. training start), sending the same request twice can create duplicate work. Use idempotency keys if the API supports them (e.g. Idempotency-Key header) when retrying.

​
Compatibility

Inference and chat completions follow an OpenAI-compatible shape. You can use the OpenAI SDK or any HTTP client by setting the base URL to the platform and the authorization header to your API key.