Seedance 2.5 Video Extend: AI Video Editor

Extend an existing video with a new cinematic continuation using Seedance 2.5 Video Extend. Try free — pay per generation.

Tutorial

📝

Overview

About this model

Seedance 2.5 Video Extend is the video-extension endpoint in the Seedance 2.5 family. It extends an input video with a new cinematic continuation generated from its last frame and a natural-language prompt, optionally interpolating toward a target end frame. Output resolution is 720p.

1Continuing a shot: Lengthen an existing clip with new camera movement or action, seamlessly picking up where it left off.
2Scripted transitions: Interpolate from a clip's last frame toward a specific target frame for a controlled scene change.
3Batch extension: Extend many short clips into longer sequences without re-generating the original footage.
4Story continuation: Add the next beat to a generated or filmed sequence using only a text description.
💰

Pricing & Value

Cost analysis

muapiapp$0.221/sec (720p)

Billed on the combined duration of the input video plus the generated continuation, at 65% of the base per-second rate.

* Competitor pricing is estimated based on similar model architectures and usage tiers.

⚙️

Technical Details

Configuration schema

Promptstring

Desired cinematic continuation — action, camera movement, lighting, mood.

Default ValueThe car keeps driving down the coastal road as the sun sets, camera slowly pulling back.
Videostring

URL of the video to extend. Generation continues from its last frame.

Default Valuehttps://cdn.muapi.ai/assets/seedance-2.5-video-extend-in.mp4
Last Imagestring

Optional target frame URL. When set, the continuation interpolates from the input video's final frame toward this image.

Default Valueundefined
Durationint

Length of the extension clip in seconds.

Default Value5
Aspect RatioEnum (7 options)

Aspect ratio of the output video.

Default Value16:9
Generate Audioboolean

Whether to generate synchronized audio for the new segment.

Default Valuetrue
Seedint

Random seed for reproducible generation. Use -1 for random.

Default Value42
📖

Implementation Guide

Developer documentation

How to Use Seedance 2.5 Video Extend

  1. Provide the Source Video

    • Pass a video URL. Generation continues from its last frame.
  2. Describe the Continuation

    • Describe the desired action, camera movement, lighting, and mood in prompt.
    • Example: "The car keeps driving down the coastal road as the sun sets, camera slowly pulling back."
  3. Optional Target Frame

    • Set last_image to interpolate from the input video's final frame toward a specific target image.
  4. Configure Output

    • Set duration (4–30s) for the length of the new segment, aspect_ratio, generate_audio, and optionally seed.
  5. Submit via the API

curl -X POST https://api.muapi.ai/api/v1/seedance-2.5-video-extend \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "The car keeps driving down the coastal road as the sun sets, camera slowly pulling back.",
    "video": "https://example.com/input.mp4",
    "duration": 5
  }'
  1. Poll and Download
    • Poll GET /api/v1/predictions/{request_id}/result until status is completed, then retrieve the video URL from output.

Common Questions

Frequently asked

What does Seedance 2.5 Video Extend do?

It generates a new cinematic continuation appended to an input video, starting from that video's last frame and guided by a text prompt.

What is `last_image` for?

An optional target frame URL. When set, the continuation interpolates from the input video's final frame toward this image instead of following the prompt alone.

How much does it cost?

Billed at 65% of the base per-second rate ($0.221/sec at 720p) on the combined duration of the input video plus the generated continuation, since a video reference is always present.

How long can the extension segment be?

Between 4 and 30 seconds, configurable via the `duration` parameter.

Does the extension include audio?

Yes by default — set `generate_audio` to `false` to skip generating synchronized audio for the new segment.