How is this different from frame interpolation between two images?

Interpolation warps existing pixels from the endpoints. Seedance 2.0 synthesizes new in-between frames using two-endpoint conditioning, enabling plausible motion and detail that aren’t present in either input while still honoring both boundaries.

Do my first and last frames need to match in size or aspect ratio?

Yes. The ByteDance2FirstLastFrameNode expects matching resolutions for stable results. Prepare both frames at the same size before loading them with LoadImage to avoid unintended rescaling or artifacts.

Can it handle large viewpoint or subject changes between the first and last frames?

It can infer substantial motion, but extreme jumps may increase artifacts or ambiguity. For best results, keep the endpoints visually related and adjust duration (more frames) to give the model room to plan a coherent path.

How do I ensure consistent, repeatable outputs?

Use a fixed seed and keep inputs constant. In ComfyUI, set or lock the seed at the graph or node level, then rerun to reproduce the same sequence deterministically.

Seedance2.0: First-Last-Frame to Video

Seedance 2.0 brings a boundary‑conditioned approach to video generation: supply the first and last frames, and the model synthesizes the in‑between motion with striking temporal stability. Instead of guessing a trajectory from a single image or relying on text prompts, the pipeline conditions on both endpoints to infer a coherent path for objects, lighting, and camera movement. In ComfyUI, this is encapsulated by the ByteDance2FirstLastFrameNode, which takes two images and produces a full sequence that faithfully lands on your final frame.

This first‑last‑frame to video (FLF2V) method reduces identity drift, flicker, and composition shifts that commonly plague single‑frame or text‑to‑video methods. Unlike basic frame interpolation—which can only warp what already exists—Seedance 2.0 synthesizes novel content between your keyframes while preserving structure and intent. The result is cleaner motion planning, stronger subject consistency, and more predictable outcomes for real productions.

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.

GET https://comfy.org/workflows/download/dc1311a65bc4.json

Fetching workflow JSON…

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

# Install (beta)
pip install comfy-sdk        # Python
npm i @comfyorg/sdk          # TypeScript

# Run "Seedance2.0: First-Last-Frame to Video" (Python)
from comfy_sdk import Comfy

client = Comfy(api_key="comfyui-...")

# This workflow, exported in API format (see note below)
wf = client.workflows.from_file("api_seedance2_0_flf2v_api.json")

asset = client.assets.from_file("input.png")
wf.set_input("3", "image", asset)  # LoadImage

job = client.run(wf)
for output in job.get_outputs("2"):  # SaveVideo
    output.to_file(output.name)

The SDK takes a workflow in API format: open this workflow in ComfyUI and use File → Export Workflow (API). API access requires a Comfy Cloud plan with an API key. SDK docs · Get an API key

FAQ

Frequently Asked Questions

View all workflows