Skip to main content
ModelRunner
bytedance avatar

Seedance 2.5 First & Last Frame API

bytedance/seedance-v2.5/first-last-frame

Generate a video that starts on one image and ends on another — pin both ends of the shot and get a single take up to 30 seconds long with synchronized sound.

Priced by resolution
OpenAPI

Model Input

Input

Describe what happens BETWEEN the two supplied frames: the action, the camera move, and the sound. The opening and closing frames are already fixed by the images, so spend the prompt on the transit between them rather than on re-describing either end.

Still image used as the FIRST frame of the clip. 300-6000 px on a side, under 30 MB, aspect ratio between 1:2.5 and 2.5:1. JPEG, PNG, WebP, BMP, TIFF, GIF, HEIC and HEIF are accepted. The finished clip takes its shape from the supplied frames.

Still image used as the LAST frame of the clip. Same formats and size limits as the first frame. Use a frame that shares the first frame's scene and subject, and give it the same aspect ratio: the model generates the transit between the two images, it does not cut between unrelated shots.

Output resolution of the clip. 720p costs more per second of video than 480p.

Clip length in seconds, from 4 to 30. Leave at -1 (the default) to let the model choose an appropriate whole-second length inside that range. Billing is per second of finished video, so -1 makes the cost of a run variable.

Additional Settings

Customize your input with more control.

Generate a synchronized soundtrack (dialogue, ambience and sound effects) together with the picture. Set false for a silent clip; the price is the same either way.

Frame shape of the clip. adaptive is the only value this model accepts: the clip always inherits the aspect ratio of the supplied frames, so a vertical pair produces a vertical clip. Crop both images before you send them if you need a different shape.

You need to be logged in to run this model and view results.
Log in

Model Output

Output

Loading
Generated in 432 seconds
Logs (1 lines)

Model Example Requests

Examples

Example output 1

Model Pricing

Pricing

Model pricing varies by the target resolution of your output video.

480p
$0.154
per second of output video
or around 6 seconds for $1
720p
$0.347
per second of output video
or around 3 seconds for $1

Seedance 2.5 First & Last Frame API

Seedance 2.5 First & Last Frame is a image-to-video AI model by bytedance. On ModelRunner it runs through a REST API or via MCP from any AI assistant, at $0.347 per second of video.

POST https://queue.modelrunner.run/bytedance/seedance-v2.5/first-last-frame

cURL

# Submit a request to the queue. Input fields go at the top level of the
# body. The optional reserved "metadata" object holds your own flat string
# tags — stored on the request, never sent to the model; filter later with
# GET https://queue.modelrunner.run/requests?metadata=<url-encoded JSON>.
curl -X POST https://queue.modelrunner.run/bytedance/seedance-v2.5/first-last-frame \
  -H "Authorization: Key $MRUN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "image": "https://media.modelrunner.ai/pyJMS2iOt5vEEKWQo5aWD.jpeg",
    "prompt": "The moored rowboat rocks gently on its line, ripples spreading out across the still water, the mooring rope creaking …",
    "duration": 5,
    "end_image": "https://media.modelrunner.ai/90cxOiEMl16l485iLZghi.jpeg",
    "resolution": "720p",
    "aspect_ratio": "adaptive",
    "generate_audio": true,
    "metadata": {
      "project": "my-project"
    }
  }'
# → { "request_id": "...", "status_url": "...", "response_url": "..." }

# Poll status_url until "COMPLETED", then fetch the result
curl "https://queue.modelrunner.run/bytedance/seedance-v2.5/first-last-frame/requests/$REQUEST_ID/status" \
  -H "Authorization: Key $MRUN_API_KEY"
curl "https://queue.modelrunner.run/bytedance/seedance-v2.5/first-last-frame/requests/$REQUEST_ID" \
  -H "Authorization: Key $MRUN_API_KEY"

JavaScript

import { modelrunner } from "@modelrunner/client";

const result = await modelrunner.subscribe("bytedance/seedance-v2.5/first-last-frame", {
  input: {
    "image": "https://media.modelrunner.ai/pyJMS2iOt5vEEKWQo5aWD.jpeg",
    "prompt": "The moored rowboat rocks gently on its line, ripples spreading out across the still water, the mooring rope creaking …",
    "duration": 5,
    "end_image": "https://media.modelrunner.ai/90cxOiEMl16l485iLZghi.jpeg",
    "resolution": "720p",
    "aspect_ratio": "adaptive",
    "generate_audio": true
  },
});
console.log(result);

Python

import os
import requests

headers = {"Authorization": f"Key {os.environ['MRUN_API_KEY']}"}

submitted = requests.post(
    "https://queue.modelrunner.run/bytedance/seedance-v2.5/first-last-frame",
    headers=headers,
    json={
      "image": "https://media.modelrunner.ai/pyJMS2iOt5vEEKWQo5aWD.jpeg",
      "prompt": "The moored rowboat rocks gently on its line, ripples spreading out across the still water, the mooring rope creaking …",
      "duration": 5,
      "end_image": "https://media.modelrunner.ai/90cxOiEMl16l485iLZghi.jpeg",
      "resolution": "720p",
      "aspect_ratio": "adaptive",
      "generate_audio": true
    },
).json()

# Poll submitted["status_url"] until "COMPLETED", then:
result = requests.get(submitted["response_url"], headers=headers).json()

Input parameters

NameTypeRequiredDescription
promptstringyesDescribe what happens BETWEEN the two supplied frames: the action, the camera move, and the sound. The opening and closing frames are already fixed by the images, so spend the prompt on the transit between them rather than on re-describing either end.
imagestring (uri)yesStill image used as the FIRST frame of the clip. 300-6000 px on a side, under 30 MB, aspect ratio between 1:2.5 and 2.5:1. JPEG, PNG, WebP, BMP, TIFF, GIF, HEIC and HEIF are accepted. The finished clip takes its shape from the supplied frames.
end_imagestring (uri)yesStill image used as the LAST frame of the clip. Same formats and size limits as the first frame. Use a frame that shares the first frame's scene and subject, and give it the same aspect ratio: the model generates the transit between the two images, it does not cut between unrelated shots.
resolutionenumnoOutput resolution of the clip. 720p costs more per second of video than 480p. Default: "720p".
durationintegernoClip length in seconds, from 4 to 30. Leave at -1 (the default) to let the model choose an appropriate whole-second length inside that range. Billing is per second of finished video, so -1 makes the cost of a run variable. Default: -1.
generate_audiobooleannoGenerate a synchronized soundtrack (dialogue, ambience and sound effects) together with the picture. Set false for a silent clip; the price is the same either way. Default: true.
aspect_ratioenumnoFrame shape of the clip. adaptive is the only value this model accepts: the clip always inherits the aspect ratio of the supplied frames, so a vertical pair produces a vertical clip. Crop both images before you send them if you need a different shape. Default: "adaptive".

Machine-readable: OpenAPI schema · llms.txt

Use Seedance 2.5 First & Last Frame from Claude & Cursor (MCP)

Point Claude Code, Claude Desktop, Cursor, or any MCP client at the ModelRunner MCP server and Seedance 2.5 First & Last Frame becomes a tool your assistant can call directly — it authorizes via OAuth (no API key in config) and runs this model with the run_model tool using the endpoint bytedance/seedance-v2.5/first-last-frame.

MCP client config (Claude Desktop, Cursor)

{
  "mcpServers": {
    "modelrunner": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://mcp.modelrunner.run/mcp"]
    }
  }
}

Claude Code

claude mcp add --transport http modelrunner https://mcp.modelrunner.run/mcp

Then ask your assistant, for example: “Run bytedance/seedance-v2.5/first-last-frame on ModelRunner to generate video”. MCP setup guide.

Model Details

Model Details

Seedance 2.5 First & Last Frame generates a clip that begins on one image and ends on another. You supply both stills — `image` is the opening frame, `end_image` is the closing frame — and the prompt describes what happens in between: the action, the camera move, and the sound. Pinning both ends is what sets it apart: you decide where the shot lands, not only where it starts. A single take runs up to 30 seconds at 24 fps, the soundtrack — ambience, effects and dialogue — is generated with the picture rather than dubbed on afterwards and costs nothing extra, and the clip inherits the shape of the frames you supply, so a vertical pair gives a vertical clip. Output is 480p or 720p.

## Best for - Morphing one photo into another — a product before and after, a season change, a transformation - Storyboard work: pin two boards and generate the shot between them - Long interpolations — up to 30 seconds in a single continuous take - Transition and reveal shots that want ambient sound baked in, not added afterwards - Vertical or square transitions straight from the source pair, since the clip keeps its shape

## Choose another model when - You need output above 720p: this generation tops out at 720p, so use `bytedance/seedance-v2/first-last-frame`, which reaches 1080p - You have no closing frame to land on — use `bytedance/seedance-v2.5/image-to-video` - You have no source image at all and want the shot built from a prompt — use `bytedance/seedance-v2.5/text-to-video` - You need a fixed seed for reproducible runs, or a locked-off camera: neither input exists on this generation

## Tips - The two frames should share a scene and a subject — the model generates the transit between them, it does not cut between unrelated shots - Describe only the middle: the endpoints are fixed by the images, so spend the prompt on motion, camera and sound - Crop both images before sending — `aspect_ratio` only accepts `adaptive`, so the clip takes its shape from the pair - Set `duration` yourself when cost matters: billing is per second of finished video, so `-1` leaves the total open up to 30 seconds - Source images run 300-6000 px per side and under 30 MB; JPEG, PNG, WebP, HEIC and HEIF are all accepted

To run via the ModelRunner JavaScript client: ```js import { modelrunner } from "@modelrunner/client";

const result = await modelrunner.subscribe("bytedance/seedance-v2.5/first-last-frame", { input: { image: "https://media.modelrunner.ai/pyJMS2iOt5vEEKWQo5aWD.jpeg", end_image: "https://media.modelrunner.ai/90cxOiEMl16l485iLZghi.jpeg", prompt: "The fog burns off the lake as the light turns golden, the moored rowboat rocking gently on its line and ripples spreading across the water", resolution: "720p", duration: 8, generate_audio: true, }, }); ```