Available workloads

Vision Q&A

Use mlx_vlm for open-model visual question answering across images, charts, screenshots, and document images. Structured document extraction and OCR are not currently offered.

python

job = cc.submit(
    "mlx_vlm",
    {"input_uri": image_url, "prompt": "Describe the chart and identify its largest value."},
    model_id="qwen2-vl-2b",
    data_class="public",
    marketplace_execution_risk_acknowledged=True,
)