Export a trained model to GGUF, MLX, MNN, or QNN

Use this endpoint to convert a trained model into a deployment-ready format. Supported formats are GGUF (quantized), MLX (Apple Silicon), MNN (Alibaba mobile), and QNN (Qualcomm). The export runs asynchronously; poll GET /v1/exports/:id to track completion.

Method + Path

POST /v1/models/:id/export

Authentication

Include your PIN API token or JWT in the Authorization header as a Bearer token.

Authorization: Bearer <token>

Path parameters

string

required

The model identifier (e.g., md_abc123def456).

Request body

string

required

Target export format. Must be one of: gguf, mlx, mnn, qnn.

Request body example

{
  "format": "gguf"
}

Request example

curl -s -X POST "https://api.ariacompute.com/v1/models/md_abc123def456/export" \
  -H "Authorization: Bearer $PIN_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"format": "gguf"}'

Response

Returns a standard PIN success envelope with the export task.

integer

required

0 on success, nonzero on error.

object

required

Show export object

string

Unique export task identifier (e.g., ex_...).

string

The source model identifier.

string

Requested format: gguf, mlx, mnn, or qnn.

string

Task status: queued, running, succeeded, or failed.

string

Output path or file (populated after success).

string

Error message if the export failed.

string

ISO 8601 creation timestamp.

string

Human-readable message (empty on success).

Example response

{
  "code": 0,
  "data": {
    "export_id": "ex_7a8b9c0d1e2f",
    "model_id": "md_abc123def456",
    "format": "gguf",
    "status": "queued",
    "created_at": "2025-01-15T14:32:08Z"
  },
  "message": ""
}

Export process

If the model does not have a merged artifact, the system automatically merges the LoRA adapter into the base weights first.
The converter tool runs in a background goroutine:
- GGUF: Converts to FP16 then quantizes to Q4_K_M using llama.cpp.
- MLX: Runs mlx_lm.convert or the tool configured by ARIAPIN_CONVERT_MLX.
- MNN: Runs mnnconvert or the tool configured by ARIAPIN_CONVERT_MNN.
- QNN: Runs qairt-converter or the tool configured by ARIAPIN_CONVERT_QNN.
On success, the artifact path is registered and the model status returns to ready.

Errors

Code	HTTP	Meaning
401	401	Missing or invalid Bearer token.
404	404	Model not found or does not belong to you.
422	422	Unsupported format or missing converter tool.
500	500	Export executor failed (details in `data.error`).

PIN API

ROUTER API

AGENT API

Export a trained model to GGUF, MLX, MNN, or QNN

Method + Path

Authentication

Path parameters

Request body

Request body example

Request example

Response

Example response

Export process

Errors

​Method + Path

​Authentication

​Path parameters

​Request body

​Request body example

​Request example

​Response

​Example response

​Export process

​Errors

Method + Path

Authentication

Path parameters

Request body

Request body example

Request example

Response

Example response

Export process

Errors