What does the video_editing toggle do?

It switches the model into edit mode so the output follows your source video’s composition, motion, aspect, and timing. Your text instruction is applied as a guided change rather than generating a completely new video.

Will the output keep the same duration and aspect ratio?

Yes. In video_editing mode, the workflow inherits the input video’s aspect and duration, so timing and framing remain intact while the edit is applied.

How can I reduce flicker or inconsistencies across frames?

Keep instructions concise and unambiguous, work on shorter slices with Video Slice (avoid hard cuts within the same run), and iterate in small steps. If the node exposes quality/consistency controls, prefer defaults or moderate values rather than extreme settings.

Does the workflow preserve audio?

SaveVideo typically outputs a silent clip in ComfyUI. If you need audio, export the edited video and reattach the original soundtrack in a video editor.

Any tips on performance and limits?

Because ByteDance2ReferenceNode uses an API, processing time depends on clip length, resolution, and network speed. For faster, more stable edits, start with 5–15 second segments at 720p–1080p. If you encounter errors or long waits, reduce duration or resolution.

Seedance2.5: Video Editing - ComfyUI Workflow

Seedance2.5: Video Editing is a text-driven video-to-video editor built in ComfyUI. It loads your source clip with LoadVideo, optionally trims it with Video Slice, and sends the frames to ByteDance2ReferenceNode running Seedance 2.5 in video_editing mode. You describe the change in plain language—such as “replace the background with a minimalist white studio” or “remove the product label and keep everything else the same”—and the model returns a single, temporally consistent edited clip.

Technically, ByteDance2ReferenceNode handles the model/API call and enforces frame-to-frame coherence so objects, lighting, and motion remain stable as the edit carries across time. The workflow inherits the aspect ratio and duration from your input (per the video_editing mode behavior and author notes), preserving original composition and timing while applying local object removal, background swaps, scene replacements, or style transfers. MarkdownNote provides in-graph guidance, and SaveVideo assembles the processed frames into your final output video.

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.

8d3a504fd30d.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 this workflow (TypeScript)
import { Comfy } from "@comfyorg/sdk";

const client = new Comfy({ apiKey: "comfyui-..." });
const wf = await client.workflows.fromFile("workflow_api.json");
const job = await client.run(wf);
await job.getOutputs("<output-node-id>")[0].toFile("output.png");

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

Frequently Asked Questions