Qwen Image 2 Pro Edit (fal.ai)

The edit half of the Qwen Image 2 Pro pair. Where FAL_QwenImage2Pro generates, FAL_QwenImage2ProEdit takes up to three images plus an instruction and sends them to fal's fal-ai/qwen-image-2/pro/edit endpoint. Same architecture as the standard Qwen Image 2 editor - the 7B unified model that's only available through APIs - but in the higher-priced Pro lane that fal bills for the more demanding version of the same family.

This is the node you reach for when you want the new Qwen's instruction-following and 2K output on an actual edit job, and you'd rather pay for a call than host a 7B model that (as of writing) has no public weights anyway. The input side is identical to the standard edit node; the difference is the tier.

How it works

Your connected image tensors are uploaded and sent as image_urls - first image is the subject, image2/image3 are optional references. Your prompt drives the edit, the endpoint regenerates the frame, and the result comes back as an IMAGE tensor ready to preview or save.

The inputs that matter:

• image (required) and prompt (required) - the subject photo and the edit command.
• image2 / image3 - optional reference images, up to three total.
• num_images - up to 6, so you can ask for a spread of edit variations in one call (each one billed).
• enable_prompt_expansion (default true) - the model rewrites your instruction internally; turn it off for literal obedience of a precise prompt.
• image_size presets (square_hd default) or custom (512–2048px), negative_prompt, enable_safety_checker (default on), seed, output_format, private_mode.

One IMAGE output, one image per input port (first frame of each tensor - not a batch processor).

Install

cd ComfyUI/custom_nodes
git clone https://github.com/elle-trudgett/fal-ai-comfyui-nodes
pip install -r fal-ai-comfyui-nodes/requirements.txt

or ComfyUI Manager → ComfyUI fal.ai Nodes. No model downloads. Set your key - export FAL_KEY="..." or a config.ini with [API] / FAL_KEY in the node folder for ComfyUI Desktop - from fal.ai/dashboard/keys, then restart.

Gotchas

• "FAL_KEY not set" - env var or config.ini, restart, done.
• You're paying the Pro tax on every upload and generation. Three images at Pro tier, times six variations, is a real charge. Prototype on the standard tier.
• Full-frame re-emission. Like every generative editor, pixels you didn't mention can drift; this is an instruction follower, not a surgical tool.
• No steps or guidance knobs - prompt and expansion are the tuning surface for this architecture.
• Errors surface as fal.ai error: ... with fal's message intact.

If you've already got the standard Qwen Image 2 Edit producing what you need, skip this - the Pro lane is for when the base tier comes up short on a specific job and the price is worth it. Otherwise it's the cleanest way to try Qwen's new editor before any weights ever materialize.