API endpoints

Main REST endpoints by area.

API endpoints

All endpoints are relative to the base URL https://api.llmtune.io/v1 for inference/agent or https://api.llmtune.io for fine-tuning. Send Authorization: Bearer sk_... on every request.

Inference

Method	Path	Description
POST	/v1/models/{modelId}/inference	Single-prompt completion. Body: prompt, temperature, maxTokens, etc.
POST	/v1/chat/completions	Chat completions. Body: model, messages, temperature, max_tokens, optional stream.
POST	/v1/confidential/chat/completions	Confidential chat completions with enhanced security. Body: model, messages, temperature, max_tokens, optional stream.

Agent

Method	Path	Description
GET	/api/agent/v1/models	List models available for the Agent.
POST	/api/agent/v1/chat/completions	Chat with optional tools. Body: model, messages, tools, tool_choice, etc.

Fine-tuning

Method	Path	Description
POST	/api/fine-tune/training/start	Start a fine-tuning job. Body: base model, dataset, hyperparameters, optional webhook.
GET	/api/training/{jobId}	Get job status, progress, and metrics.
POST	/api/training/{jobId}/cancel	Cancel a running job (if supported).

Exact paths and request/response schemas may vary slightly; use the dashboard and the OpenAPI spec (if available) for the current contract. Error responses use standard HTTP status codes and a JSON error (and optional message) field.