Do I need a reference image, or can I generate from text only?

References are optional. LoadImage provides identity and style cues (faces, products, environments) for image-conditioned results. If you omit references, ByteDance2ReferenceNode will generate from the prompt alone.

How do I make the camera motion feel intentional instead of random?

Write concise, directive prompts. Use specific camera verbs (dolly-in, push, crane down, handheld sway, static lock-off), define beats (start, middle, end), and explicitly instruct the final frame (e.g., end on a locked close-up). GeminiNode helps structure this into Seedance 2.0 format, but clarity in your brief matters most.

What settings control duration, resolution, or aspect ratio?

These are configured on ByteDance2ReferenceNode, subject to what your node version exposes and what the backend supports. Set aspect ratio and duration there if available. Keep durations short for crisper motion and stronger final frames.

How do I get consistent subjects across multiple shots?

Reuse the same reference images and keep your key identity cues and framing language consistent across prompts. Avoid over-describing style changes between shots, and maintain similar camera and composition cues for continuity.

Seedance 2.0 Cinematic Prompt Studio

Seedance 2.0 Cinematic Prompt Studio is a ComfyUI workflow for turning concise, director-style prompts into production-ready cinematic shots. It pairs optional image references with a structured Seedance 2.0 prompt that emphasizes intentional camera motion, visual storytelling beats, and a strong final frame. The workflow is organized into three groups: Load Images, Generate Seedance 2.0 Prompt, and Generate Video.

Technically, LoadImage supplies one or more visual references, while GeminiNode converts a short shot brief into a Seedance 2.0–formatted prompt (camera verbs, lens/composition cues, lighting, pacing, and a closing frame directive). These are fed into ByteDance2ReferenceNode, which performs image-conditioned or text-only video generation via the Seedance 2.0 backend. PreviewAny lets you inspect intermediate outputs and play back results in-canvas, and SaveVideo writes the final render to disk. A MarkdownNote in the graph provides on-canvas guidance and a prompt template so you can iterate quickly and keep prompts concise and cinematic.

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.

68c39b4c6a6f.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