Black Forest Labs Flux 3 Video Edit API Documentation
Playground
Try it on WaveSpeedAI!FLUX 3 Video Edit applies prompt-guided changes to existing videos while preserving motion, timing, and framing, supporting precise video modification, visual style updates, scene refinements, creative edits, ads, and production workflows. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Features
FLUX 3 Video Edit makes targeted changes to an existing video using natural-language instructions. Change an object’s appearance, adjust the environment, or restyle a scene while preserving the source motion, timing, and framing.
Why Choose This?
-
Prompt-guided video editing
Edit an existing video by describing the changes you want in natural language. -
Motion and timing preservation
Keep the source video’s motion, pacing, camera movement, and framing while applying visual changes. -
Targeted visual changes
Modify objects, environments, style, lighting, or scene details without generating a new clip from scratch. -
Simple 720p workflow
Upload a short MP4 video, write an edit prompt, and receive a 720p edited video. -
Focused short-clip editing
Designed for short video edits where the original timing and structure should remain consistent.
Parameters
| Parameter | Required | Description |
|---|---|---|
| video | Yes | Source MP4 video to edit. The input must be under 50 MB and under 15 seconds. |
| prompt | Yes | Natural-language edit instruction describing the changes to apply. Maximum length: 4096 characters. |
The output is a 720p edited video. This endpoint does not expose duration, resolution, aspect ratio, audio-generation, or seed controls. Output timing follows the source clip.
How to Use
- Upload a source video — Provide a short MP4 video under
50 MBand under15seconds. - Write an edit prompt — Describe the specific changes you want to apply.
- Submit — Run the video edit request.
- Retrieve the output — Get the edited video URL from the prediction outputs.
Pricing
Pricing is based on input video duration.
The price is $0.03 per billable second for 720p video.
| Input Duration | Billable Seconds | Cost |
|---|---|---|
| 1s | 1 | $0.03 |
| 5s | 5 | $0.15 |
| 5.1s | 6 | $0.18 |
| 10s | 10 | $0.30 |
| 14.5s | 15 | $0.45 |
prompt does not add a separate charge. Billing is based on input duration, not a post-generation measurement of the output.
Best Use Cases
- Object appearance edits — Change the look, material, color, or style of objects in a video.
- Environment changes — Adjust backgrounds, locations, atmosphere, or scene details.
- Style transfer — Restyle a clip while preserving its original motion and timing.
- Short-form creative edits — Modify short clips for social media, ads, concepts, or campaign assets.
- Prompt-based iteration — Test different edit directions on the same source video.
Pro Tips
- Describe what should change and what should stay the same.
- Use clear source footage with the motion and framing you want to preserve.
- Keep the prompt focused on one main edit direction for more stable results.
- Mention lighting, background, object appearance, and style when they matter.
- Use short clips for faster iteration and easier review.
- This endpoint edits a supplied clip; it does not extend the duration or expose output resolution controls.
Notes
videoandpromptare required.- The source video must be an MP4 file under
50 MBand under15seconds. - The output is a 720p edited video.
- Output timing follows the source clip.
- The endpoint does not expose duration, resolution, aspect ratio, audio-generation, or seed controls.
- The billing cap does not make videos above the input duration limit supported.
Related Models
- FLUX 3 Image-to-Video — Animate a reference image into a high-quality video.
- FLUX 3 Text-to-Video — Generate video directly from a text prompt.
- FLUX 3 Draft Image-to-Video — Create fast draft videos from a reference image for motion testing.
- FLUX 3 Draft Text-to-Video — Generate fast draft videos directly from text prompts.
Authentication
For authentication details, please refer to the Authentication Guide.
API Endpoints
Submit Task & Query Result
set -euo pipefail
export WAVESPEED_API_KEY="your-api-key"
REQUEST_BODY=$(cat <<'JSON'
{
"video": "https://interactive-examples.mdn.mozilla.net/media/cc0-videos/flower.mp4",
"prompt": "A cinematic ocean wave at sunrise, highly detailed"
}
JSON
)
# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
-X POST "https://api.wavespeed.ai/api/v3/black-forest-labs/flux-3/video-edit" \
-H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
-H "Content-Type: application/json" \
-d "${REQUEST_BODY}")
TASK=$(printf '%s' "${SUBMIT_RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "${TASK}" | jq -r '.id // empty')
if [ -z "${PREDICTION_ID}" ]; then
printf 'Submission response did not contain a prediction id
' >&2
exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"
# 2. Poll until the prediction finishes.
while true; do
RESPONSE=$(curl --silent --show-error --fail-with-body \
"${RESULT_URL}" \
-H "Authorization: Bearer ${WAVESPEED_API_KEY}")
RESULT=$(printf '%s' "${RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')
case "${STATUS}" in
completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
failed|cancelled|timeout|deleted) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
*) sleep 2 ;;
esac
doneParameters
Task Submission Parameters
Request Parameters
| Parameter | Type | Required | Default | Range | Description |
|---|---|---|---|---|---|
| video | string | Yes | - | URL of the input MP4 video. The video must be under 50 MB and under 15 seconds. | |
| prompt | string | Yes | - | Describe how to edit the input video. The clip is re-rendered while preserving its motion, timing, and framing. Maximum 4096 characters. |
Response Parameters
| Parameter | Type | Description |
|---|---|---|
| code | integer | HTTP status code (e.g., 200 for success) |
| message | string | Status message (e.g., “success”) |
| data.id | string | Unique identifier for the prediction, Task Id |
| data.model | string | Model ID used for the prediction |
| data.outputs | array | Output values, usually URL strings; some models return text strings or structured result objects (empty when status is not completed) |
| data.urls | object | Object containing related API endpoints |
| data.status | string | Task status. completed is successful; failed, cancelled, timeout, and deleted are failure terminal statuses. |
| data.created_at | string | ISO timestamp of when the request was created (e.g., “2023-04-01T12:34:56.789Z”) |
| data.error | string | Error message (empty if no error occurred) |
| data.timings | object | Object containing timing details |
| data.timings.inference | integer | Inference time in milliseconds |
Result Request Parameters
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
| id | string | Yes | - | Task ID |
Result Response Parameters
| Parameter | Type | Description |
|---|---|---|
| code | integer | HTTP status code (e.g., 200 for success) |
| message | string | Status message (e.g., “success”) |
| data | object | The prediction data object containing all details |
| data.id | string | Unique identifier for the prediction |
| data.model | string | Model ID used for the prediction |
| data.outputs | array<string | object> | Array of generated outputs (empty when status is not completed). Items are usually URL strings, but may be text strings or structured result objects, depending on the model. |
| data.urls | object | Object containing related API endpoints |
| data.status | string | Status: completed is successful; failed, cancelled, timeout, and deleted are failure terminal statuses |
| data.created_at | string | ISO timestamp of when the request was created |
| data.error | string | Error message (empty if no error occurred) |
| data.timings | object | Object containing timing details |
| data.timings.inference | integer | Inference time in milliseconds |