API Endpoints
This page summarizes the primary public REST endpoints exposed by LLMTune’s API. All paths below are relative to the base URL:Inference
Run completions against hosted or deployed models.| Method | Endpoint | Description |
|---|---|---|
| POST | /models/{modelId}/inference | Run inference against a deployed model. |
| POST | /playground/inference | Mirror of the UI playground for quick tests. |
| POST | /batch/inference | Submit a batch of up to 100 inference requests. |
Example – Single Inference
Fine-Tuning
Launch and monitor fine-tuning jobs.| Method | Endpoint | Description |
|---|---|---|
| POST | /fine-tune | Submit a fine-tune job. |
| GET | /fine-tune/{jobId} | Fetch job status, metrics, and errors. |
Example – Submit Fine-Tune
Webhooks
Receive lifecycle events for fine-tunes and deployments. LLMTune sends signed POST requests to your configured webhook URL with payloads like:training.startedtraining.completedtraining.failedmodel.deployed
Error Handling
All endpoints return a standard error schema on failure:401– Invalid or missing API key402– Insufficient credits404– Model or job not found429– Rate limited500– Unexpected server error