Google Veo2 Image-to-Video (fal)

Veo 2 is Google's video model - the one people used for those uncannily smooth vlog-style clips before Veo 3 added audio. Its strength is physics and realism: motion that looks physically plausible instead of the melty warping cheaper models produce. This node animates a still image with Veo 2 through fal.ai, from a ComfyUI graph.

Hosted call, per-clip billing, nothing local. Veo is closed and Google-only otherwise, so fal is a rare way to script it into a graph.

How it works

You give it a starting image and a prompt describing the motion, and Veo 2 generates a short clip that begins from your frame. The image fixes the look; the prompt drives what moves. Veo's edge over open image-to-video models is coherence - it holds subjects together through motion better than most, which is why it's a go-to when you need the result to look real rather than dreamlike.

Note this is Veo 2, not 3 - no native audio here. Audio is Veo 3's headline feature (and there's a separate Veo 3.1 node in this pack for it). Veo 2 gives you silent, high-quality motion.

The inputs and outputs that matter

• prompt (required, multiline) - describe the motion and camera, not the image. The still already carries the scene.
• image (required, IMAGE) - the first frame. Composition, subject, and style all come from here, so use a clean, high-quality still.
• aspect_ratio - auto (default, matches your image), auto_prefer_portrait, 16:9, or 9:16. Leave it on auto unless you specifically need a different crop.
• duration - 5s, 6s, 7s, or 8s. Longer costs more; 8s is the ceiling here.

Output is a STRING - a URL to the video on fal, not a previewable VIDEO. Run it through the pack's Fal Save Media from URL node to download, or into a downstream fal node's *_direct_url. fal URLs expire; save what you want.

How to install it

Ships in gokayfem's ComfyUI-fal-API pack. ComfyUI Manager: search ComfyUI-fal-API, install, restart. Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/gokayfem/ComfyUI-fal-API.git
pip install -r ComfyUI-fal-API/requirements.txt

Add your fal key (copy config.ini.example → config.ini, paste from fal.ai/dashboard/keys, or export FAL_KEY=your_key), restart.

Common issues & troubleshooting

No sound in the output. Correct - Veo 2 doesn't generate audio. If you want synced audio with your video, use the Veo 3.1 node in this pack instead.

Motion ignores the prompt. Keep the prompt about movement and camera; don't restate what's already visible in the image. Redundant scene description dilutes the motion instruction.

The clip drifts from your image. A sharper, less ambiguous input frame holds far better than a busy or low-res one. Image-to-video always trades some fidelity for movement.

Output won't preview. It's a URL string - download it with Fal Save Media from URL first.

Insufficient funds / credit lag. fal charges per call and is a legit platform, but top-ups have occasionally been slow to post. Let credits settle rather than repaying. Google's content policy is enforced on the fal side; a policy refusal comes through as readable error text in ComfyUI.