sgl-project/sglang · error · HTTPException

Invalid request body: {e}

Error message

Invalid request body: {e}

What it means

Thrown when constructing MeshGenerationsRequest from the request body fails, or when saving an inline/base64 image source raised. Any exception inside the body-parsing block (including image saving) is converted to HTTP 400 'Invalid request body'.

Source

Thrown at python/sglang/multimodal_gen/runtime/entrypoints/openai/mesh_api.py:205

        try:
            body = await request.json()
        except Exception:
            body = {}
        try:
            payload: Dict[str, Any] = dict(body or {})

            if payload.get("input_image"):
                img_src = payload.pop("input_image")
                uploads_dir = os.path.join("outputs", "uploads")
                os.makedirs(uploads_dir, exist_ok=True)
                input_path = await save_image_to_path(
                    img_src,
                    os.path.join(uploads_dir, f"{request_id}_input_image"),
                )

            req = MeshGenerationsRequest(**payload)
        except Exception as e:
            raise HTTPException(status_code=400, detail=f"Invalid request body: {e}")

    if not input_path:
        raise HTTPException(
            status_code=422,
            detail="An input image is required for mesh generation",
        )

    sampling_params = _build_sampling_params_from_request(request_id, req, input_path)
    job = _mesh_job_from_sampling(request_id, req, sampling_params)
    await MESH_STORE.upsert(request_id, job)

    batch = prepare_request(
        server_args=server_args,
        sampling_params=sampling_params,
    )

    asyncio.create_task(_dispatch_job_async(request_id, batch))
    return MeshResponse(**job)

View on GitHub (pinned to 0132848349)

Solutions

1. Validate the request body against the MeshGenerationsRequest schema (required fields, types) before sending
2. Check the endpoint's request model definition for required vs optional fields
3. Ensure the image field (file or source) is provided in the accepted form

Example fix

// before
{"model": "mesh-model", "prompt": 123}
// after
{"model": "mesh-model", "prompt": "generate 3d mesh"}

# client-side: check required fields before POST
assert isinstance(body.get('model'), str) and body.get('model')
assert isinstance(body.get('prompt', 'generate 3d mesh'), str)

if resp.status_code == 400 and 'Invalid request body' in detail:
    log(resp.text); fix schema per MeshGenerationsRequest

When it happens

Trigger: POST /mesh/generations with a JSON body missing required fields, wrong field types, unknown keys rejected by the Pydantic model, or a malformed image source reference.

Common situations: Sending extra parameters the model doesn't define; passing a non-string model id; typo'd field names like 'promt'; using an API version whose request schema changed.

Related errors

• Failed to process image source: {str(e)}
• An input image is required for mesh generation
• Mesh not found
• Mesh has been uploaded to cloud storage. Please use the clou
• Generation is still in-progress

AI-assisted analysis of sgl-project/sglang@0132848349 (2026-08-28). Data as JSON: /api/errors/d5a7e08ab80bca52. Report an issue: GitHub.