How do I keep visuals stable across frames?

Use a fixed seed, keep the subject and setting clearly defined in the prompt, and avoid conflicting motion cues. Start with a shorter duration to validate stability before rendering longer clips.

How can I control camera motion?

Describe the motion in the prompt (e.g., “static shot,” “slow dolly-in,” “smooth pan left”). Clear, simple motion language tends to produce more predictable results than mixing multiple complex moves.

What resolution and fps should I start with?

Begin at a modest resolution (e.g., 512–720p range) and 12–24 fps to iterate quickly. Once satisfied, scale up resolution or fps. Higher settings increase VRAM usage and render time.

Does the workflow generate or embed audio?

No. This setup uses ByteDance2TextToVideoNode for visuals and SaveVideo for encoding. Export the video and add music or sound effects in your video editor for precise sync.

Seedance 2.0: Text to Video - ComfyUI Workflow

Seedance 2.0: Text to Video turns plain-language prompts into cinematic video clips inside ComfyUI. The workflow is built around ByteDance2TextToVideoNode running the Seedance 2.0 model to synthesize a coherent sequence of frames directly from text. You control duration (frame count), playback speed (fps), resolution, and seed, which together determine the look, timing, and repeatability of the output. SaveVideo then encodes the generated frames into a single video file (e.g., MP4/WebM) at your chosen frame rate.

Technically, ByteDance2TextToVideoNode conditions a video diffusion process on your prompt and iteratively produces a temporally consistent stack of images. Fixing the seed and keeping motion cues clear in the prompt helps maintain stable visuals across frames. Describing camera-like behavior in the text (e.g., “slow dolly in,” “static tripod,” “handheld”) guides motion without extra nodes. SaveVideo collects the node’s frame output and writes it to disk; if you need music or sound effects, export your video and sync audio later in a video editor.

API

Use this workflow from code

Every Comfy workflow is a JSON graph. The payload below is this workflow, exactly as ComfyUI runs it — fetch it from the URL, keep it in version control, or load it in ComfyUI and run it node by node.

api_seedance2_0_t2v.json

Fetching workflow JSON…

Run it from TypeScript or Python with the Comfy SDK. The same code targets Comfy Cloud or a ComfyUI you host yourself — only the base URL changes.

// Install (beta)
npm i @comfyorg/sdk

// Run "Seedance 2.0: Text to Video" (TypeScript)
import { Comfy } from "@comfyorg/sdk";

const client = new Comfy({ apiKey: "comfyui-..." });

// This workflow, exported in API format (see note below)
const wf = await client.workflows.fromFile("api_seedance2_0_t2v_api.json");

const job = await client.run(wf);
await job.getOutputs("344")[0].toFile("output.png");  // SaveVideo

The SDK takes a workflow in API format: open this workflow in ComfyUI and use File → Export Workflow (API).

Comfy Cloud API access requires a plan with an API key.

SDK docs Get an API key

FAQ