FLUX 3 Video Extend: AI Video Editor
Continue any video clip with prompt-guided motion and native audio using FLUX 3 Video Extend by Black Forest Labs. Try free — pay per generation.
Continue any video clip with prompt-guided motion and native audio using FLUX 3 Video Extend by Black Forest Labs. Try free — pay per generation.
About this model
FLUX 3 Video Extend continues an existing video clip with prompt-guided motion, scene development, and camera movement — with optional native synchronized audio — using Black Forest Labs' unified image/video/audio architecture. It's the way to chain FLUX 3 Text to Video or FLUX 3 Image to Video outputs into longer sequences, or continue any existing MP4 clip under 50 MB and 15 seconds. Once extended, run the result through FLUX 3 Video Upscaler for a higher-resolution final pass.
Cost analysis
| Provider | Cost | Notes |
|---|---|---|
| muapiapp | $0.25/sec (720p), $0.42/sec (1080p) | Pay-per-generation pricing, no subscription required. Audio generation included at no extra charge. |
| Fal.ai | Not available | FLUX 3 Video Extend is not yet listed on Fal.ai. |
| Replicate | Not available | FLUX 3 Video Extend is not yet listed on Replicate. |
Pay-per-generation pricing, no subscription required. Audio generation included at no extra charge.
FLUX 3 Video Extend is not yet listed on Fal.ai.
FLUX 3 Video Extend is not yet listed on Replicate.
** Competitor pricing is estimated based on similar model architectures and usage tiers.
Configuration schema
| Parameter | Type | Description | Default |
|---|---|---|---|
| Prompt | string | Text prompt describing how the video should continue. | The camera pulls back to reveal a sunrise breaking over the skyline, birds taking flight. |
| Video URL | string | URL of the source video to extend. Must be under 50 MB and under 15 seconds. | https://d3adwkbyhxyrtq.cloudfront.net/webassets/videomodels/flux-3-video.mp4 |
| Aspect Ratio | Enum (6 options) | Aspect ratio of the output video. | 16:9 |
| Duration | int | Length of the extension in seconds. | 5 |
| Resolution | Enum (2 options) | Output video resolution. | 720p |
| Generate Audio | boolean | Whether to generate synchronized native audio for the extended video. | true |
Text prompt describing how the video should continue.
The camera pulls back to reveal a sunrise breaking over the skyline, birds taking flight.URL of the source video to extend. Must be under 50 MB and under 15 seconds.
https://d3adwkbyhxyrtq.cloudfront.net/webassets/videomodels/flux-3-video.mp4Aspect ratio of the output video.
16:9Length of the extension in seconds.
5Output video resolution.
720pWhether to generate synchronized native audio for the extended video.
trueDeveloper documentation
Provide a source video — an MP4 under 50 MB and under 15 seconds, via video_url.
Write a continuation prompt — describe the next action, scene development, camera movement, mood, lighting, and visual continuity.
Configure output — set resolution (720p or 1080p) and duration (5–20 seconds) for the extension.
Toggle generate_audio for synchronized native audio on the extended portion, at no extra charge.
Submit the request:
curl -X POST https://api.muapi.ai/api/v1/flux-3-video-extend \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"prompt": "The camera pulls back to reveal a sunrise breaking over the skyline", "video_url": "https://example.com/clip.mp4", "resolution": "1080p", "duration": 5, "generate_audio": true}'
/api/v1/predictions/{request_id}/result until completed.Frequently asked
It continues an existing video clip with new prompt-guided motion and scene development, using Black Forest Labs' unified image/video/audio architecture.
The source clip must be an MP4 under 50 MB and under 15 seconds long.
Yes, synchronized native audio for the extended portion can be toggled with `generate_audio` at no extra charge.
5 to 20 seconds, billed rounded up to the next whole second.
Muapi bills $0.25/second at 720p and $0.42/second at 1080p for the extended portion, rounded up to the next whole second, capped at 20 seconds.
Yes — feed the output of one extension back in as the `video_url` of the next call to keep building a longer sequence.