Pixal3D (Pixel-Aligned 3D Generation)

Most of the BizyAir nodes do the familiar stuff - image in, image out, video in, video out. This one is the odd one out, and that's exactly why you're here: Pixal3D takes a single photo and generates a 3D asset from it, all inside your ComfyUI graph, with zero local 3D pipeline. No NeRF training, no mesh editing, no obj files cluttering your disk - the heavy lifting happens on SiliconFlow's cloud via the pixal3d-official/image-to-3d endpoint.

What you're actually getting

This is an API node in the BizyAir "3D Creation" category. You feed it an image, the cloud runs pixel-aligned 3D generation, and the node hands the result back as a native ComfyUI IMAGE - rendered views of the generated 3D, ready to preview or save - alongside the usual request_id and urls outputs for tracing and direct file access. It's the pragmatic ComfyUI answer to "I need a rough 3D asset of this object without learning a 3D toolchain." Think game-asset prototyping, product mockups, or reference geometry, not production-grade retopology.

Inputs that matter

Honestly, there aren't many, which is refreshing:

• images - the source photo (the tooltip just calls it "Photo"). A clean, well-lit, single-subject shot with the object filling the frame gives the best result; busy backgrounds confuse pixel-aligned 3D.
• resolution - a text field that defaults to 1024, the target resolution for the generation. Bump it for more detail, keep it low for quick tests that don't eat credits.
• api_config - per-request base_url/api_key override; ignore it unless you're testing alternate endpoints.
• skip_error - flip on and a failed request returns a placeholder instead of halting your workflow.

Outputs: image (wire into a preview or save node), request_id (for support), and urls (direct links to the result files).

Install and get going

Install the pack via ComfyUI Manager (search "BizyAirPlus") or:

cd /path/to/ComfyUI/custom_nodes
git clone https://github.com/siliconflow/BizyAirPlus.git
cd BizyAirPlus
python -m pip install -r requirements.txt

Restart ComfyUI, then create an account at bizyair.ai, grab an API key, and set it under Settings > BizyAirPlus > API Key (or as the BIZYAIR_API_KEY environment variable). Then click the BizyAirPlus button in the action bar until it shows ON - like every BizyTRD node, this one does nothing while that toggle is off.

Gotchas

The standard BizyAir suspects apply: missing key, cloud toggle off, or the pack installed into the wrong Python (python -m pip show bizyair-cloudberry to verify; check the ComfyUI console for the real error). The 3D-specific gotcha is input hygiene - a busy, low-contrast photo will produce a lumpy, unconvincing asset, and since this is a paid, metered cloud service, that's a paid lumpy asset. Also set expectations: this returns images of the 3D result through the node's output, so pair it with the urls output if your downstream pipeline needs the actual asset files rather than the rendered views.