ComfyUI-emaki

ComfyUI-emaki

ComfyUI nodes for emaki (emaki.ai), the Japanese-language AI video generator for Seedance 2.5 and Seedance 2.0. Run Seedance 2.5 text to video and image to video from a graph, plus image models for the first frame. No local GPU, no model downloads: one emaki API key.

| Node | Inputs | Outputs | |---|---|---| | emaki Text to Video (Seedance 2.5) | prompt, model, duration, resolution, aspect ratio, audio | path of the saved MP4 (ComfyUI output folder), url | | emaki Image to Video (Seedance 2.5) | IMAGE (first frame), prompt, model, duration, resolution, aspect ratio, audio | path of the saved MP4, url | | emaki Generate Image | prompt, model, aspect ratio | IMAGE, url |

Chain Generate Image (or any IMAGE in your graph) into Image to Video to animate a still you control.

Install

ComfyUI Manager: search ComfyUI-emaki. Or clone into custom_nodes:

cd ComfyUI/custom_nodes
git clone https://github.com/hanshs474/ComfyUI-emaki

No extra Python dependencies: the client uses the standard library, and the nodes use the torch, numpy and Pillow that ComfyUI already ships.

API key

Sign up on emaki.ai, create a key at emaki.ai/settings/apikeys, and either paste it into the node's api_key field or set EMAKI_API_KEY before starting ComfyUI. Each run costs credits on your emaki account, the same price as on the website.

Models

• Video: seedance_2_5 (default; 4 to 30 seconds, 480p / 720p / 1080p), seedance_2 (Seedance 2.0), seedance_2_fast, seedance_2_mini, minimax_h3, veo_3_1
• Images: nano_banana_2_lite (default, the cheapest), nano_banana_2, seedream_5_lite, gpt_image_2

Image to Video uploads the frame to emaki's storage and passes its URL as the first frame; adaptive keeps the frame's own shape. Video takes minutes and the node blocks until the file is ready (15-minute limit). Change seed to force a new run with the same inputs. Every run also lands in your library on the site, where the AI video upscaler can sharpen it.

日本語での使い方は シーダンス へ。

MIT licensed. Made by emaki.