Deployment API

Prefix: /api/deployment

Source: backend/app/api/deployment.py

Start Server

POST /api/deployment/start

{
  "model_path": "/path/to/model",
  "host": "127.0.0.1",
  "port": 8080
}

Starts mlx_lm.server as a subprocess. Returns PID and status.

Stop Server

POST /api/deployment/stop

Kills the running server process.

Get Status

GET /api/deployment/status

{
  "running": true,
  "pid": 12345,
  "uptime_seconds": 3600
}

pid and uptime_seconds are null when the server is not running.

Get Logs

GET /api/deployment/logs?since=1709000000

Returns server output lines since the given Unix timestamp. Buffer capped at 500 entries.

{
  "logs": [
    { "timestamp": 1709000001, "line": "Server started on 127.0.0.1:8080" }
  ]
}