Seedance Lite Text-to-Video

The entry point of the pack's video side: a prompt in, a short clip out, no image, no video model, no local GPU. Seedance Lite text-to-video (seedance-1-0-lite-t2v-250428) is ByteDance's affordable cloud video model, and this node is the ComfyUI wrapper around it. If you've ever wanted to test whether a video idea has legs before committing to a heavier local pipeline - or you just don't have the hardware to run a video model at all - this is the node that answers "what would this look like in motion?" in about a minute.

Seedance is one of the closed, API-only models the community benchmarks open video models against (the KB's ByteDance panel counts Seedance as a flagship product that stays on the API - it will never be a downloadable checkpoint). So this is pure cloud: the node builds your prompt into a text payload, creates an async generation task on BytePlus, and polls every five seconds until your video URL is ready. "Lite" is the honest qualifier: quality is solid for prompt-driven scenes, but if you're doing character work or need top fidelity, the Pro nodes or the image-driven Lite nodes will serve you better.

Inputs and outputs that matter

• prompt - the whole movie in a sentence. Say what's happening, the motion, the camera, the mood. "A lighthouse beam sweeping across fog, slow dolly, moody blue" beats "a lighthouse."
• resolution - 480p, 720p, or 1080p (the text-to-video Lite node gets the full range).
• ratio - 16:9 default, plus 4:3, 1:1, 3:4, 9:16, 21:9, and adaptive.
• duration - 3–12 seconds. Longer is slower, pricier, and more prone to drift.
• camerafixed - true locks the camera down; leave false for movement.
• seed - reproducibility, -1 for random. framepersecond is fixed at 24.

Output is a single STRING - the hosted video URL. Wire it to a preview/save node, or into Video URL to Frames to extract frames for further processing.

Install

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

Restart, and set your BytePlus API key in config.ini (under [API]) or as BYTEPLUS_API_KEY. Manager: search "ComfyUI Seed API Integration".

Where people get burned

First, the output is a URL, not a file - you need a video node to actually preview or save it, and the clip lives on BytePlus's servers. Second, an empty string back means the node swallowed an error; the console holds the truth, usually an auth, quota, or model-permission problem. On the craft side: 1080p costs more and takes longer, and text-to-video quality degrades as you ask for too much per clip - one clear action per 5 seconds works far better than a five-act script. If your result looks "off," fix the prompt before you blame the seed. And remember this is paid per render - Lite is the cheap tier, but iteration is still metered, so count your takes.