API Reference

Start the server with routeframe serve (default: localhost:11435)

POST /api/forecast

Request

{
  "model": "toto",
  "input": [[20, 22, 19, 23, 21, 25, 24, 26, 27, 25]],
  "prediction_length": 4,
  "future_exogenous": [[0, 1, 1, 0]]
}

Field	Type	Required	Description
model	string	Yes	Model name
input	float[][]	Yes	Array of time series (one per variate)
prediction_length	int	No	Future steps to predict (default: 1)
future_exogenous	float[][]	No	Known future covariate values

Response

{
  "model": "toto",
  "mean": [[21.53, 21.51, 21.52, 21.50]],
  "inference_time_ms": 4
}

GET /api/tags

List all downloaded models.

{
  "models": [
    {"name": "toto", "architecture": "toto", "size_bytes": 302996800}
  ]
}

POST /api/pull

{"model": "toto", "variant": "f16"}

POST /api/finetune

{
  "output_model": "my-model",
  "data_path": "/path/to/data.csv",
  "targets": ["cpu_usage"],
  "exogenous": ["is_holiday"],
  "steps": 1400
}

Returns {"job_id": "ft-1", "status": "running"}. Check progress with GET /api/finetune/status?id=ft-1.

POST /api/show

{"model": "toto"}

DELETE /api/delete

{"model": "toto"}

GET /api/health

{"status": "ok", "version": "0.1.0"}