Models API

List, download, and manage models via REST.

Endpoints

Method	Path	Description
GET	/api/models	Full model catalog
GET	/api/models/downloaded	Models present on disk
GET	/api/models/downloads/active	In-progress downloads
POST	/api/models/download-repo	Start a download by HF repo ID
POST	/api/models/download-cancel	Cancel an in-progress download
DELETE	/api/models/{model_id}	Delete a downloaded model
POST	/api/engine/set-model	Switch active model + restart engine

Download a model

curl -X POST http://localhost:3000/api/models/download-repo \
  -H "Content-Type: application/json" \
  -d '{"hf_repo": "Qwen/Qwen2.5-7B-Instruct"}'

Returns {"job_id": "...", "status": "downloading"}.

Cancel a download

curl -X POST http://localhost:3000/api/models/download-cancel \
  -d '{"job_id": "<job_id>"}'

Stops the download and cleans up partial files.

Switch active model

curl -X POST http://localhost:3000/api/engine/set-model \
  -d '{"model": "Qwen/Qwen2.5-7B-Instruct"}'

Saves to config and restarts the vLLM engine.