About MiniMax H3 Edit
The MiniMax H3 Edit API runs one endpoint, minimax/h3-edit/video-edit. You send an existing clip as video_url and a prompt that describes the change, and the response is the same shot with that change applied and, by default, a fresh soundtrack. It suits apps that let users restyle footage, such as turning a daytime clip into night, because the people and their motion come from the source file.
Call it with the model name minimax/h3-edit/video-edit on POST /jobs/createTask, from $0.15.
Introduction
MiniMax H3 Video edit takes a clip you already have, sent as video_url, and a prompt that describes a change. It returns a new version of the same shot: 3 to 15 seconds at 480p, 540p or 768p, with a new soundtrack generated by default. Your footage keeps driving who is in the frame, the composition and the motion; the prompt rewrites the lighting, weather, style, environment or specific elements. Up to nine reference images and up to three reference audio clips can guide identity, style or sound.
Choose a MiniMax H3 mode by starting material
| What you have at the start | Mode |
|---|---|
| a written scene or script | MiniMax H3 Text to video |
| an approved opening image | MiniMax H3 Image to video |
| one to nine reference images | MiniMax H3 Reference to video |
| a clip that needs to carry on | MiniMax H3 Video extend |
| a clip that needs to look different | MiniMax H3 Video edit |
Key features
Your clip keeps driving the people and the motion
The source decides who is in the shot, how it is framed and how everyone moves. The prompt decides what the world around that looks like. That split is what separates an edit from generating a new shot: you do not describe the scene again, only the change.
A sunny terrace turned into a storm
Our test run used a five-second 1280x720 clip from our OmniHuman 2.0 examples: a woman in a cream satin trench on a sunny seaside terrace. The prompt asked for a stormy dusk with dark clouds, rain, wet wind-blown hair and coat, and distant lightning, and ended with "Keep her face, pose and movement." At 768p with duration_seconds 5 and sound on, the result was a 1344x768 clip at 24 fps, about 5.2 seconds long, with new sound, after about three minutes of processing.
What changed: the sunny sky became a dark, cloudy dusk with a lightning flash over the sea, and the stone looked wet. What stayed: her face, her coat and her smile. Two things to expect honestly. The rain itself was light at 768p; a 480p trial of the same prompt showed stronger rain streaks. And she lifted a hand to her wind-blown hair, a gesture the source clip does not have. The prompt asked for wind-blown hair, so describing movement can bring small motion changes with it.
References are named by number
reference_image_urls takes up to nine images, called <Picture 1> to <Picture 9> in the prompt, and reference_audio_urls takes up to three audio clips, called <Audio 1> to <Audio 3>. Point an image at a look or an identity, and an audio clip at the sound you want. We have not run this endpoint with references yet, so treat your first take with them as a test: start with one image or one clip and see what it changes.
Sound, length and shape
generate_audio is on by default, which gives the edit a new soundtrack. Switch it off to keep the source clip's own audio track. duration_seconds runs from 3 to 15 and defaults to 5, so set it to the length of your source to edit the whole clip. Leave aspect_ratio out and the edit keeps the shape of the source; set it to 16:9, 9:16, 1:1, 4:3, 3:4, 21:9 or 9:21 to reframe.
Three resolutions
768p is the default and H3's own canvas. We ran 768p and a 480p trial; we have not tried 540p.
Request parameters
| Parameter | Required | Description |
|---|---|---|
aspect_ratio | No | The shape of the frame: wide, square or tall. 16:9 · 9:16 · 1:1 · 4:3 · 3:4 · 21:9 · 9:21 |
duration_seconds | No | How long the clip runs. You pay per second. ≥ 3 · ≤ 15 · Default: 5 |
generate_audio | No | Adds sound made to match the picture: voices, ambience and effects. Default: true |
prompt | Yes | The change to make: lighting, weather, style, setting or specific elements. Name references as <Picture 1> or <Audio 1>. |
reference_audio_urls | No | Audio to match: a voice, music or a rhythm. |
reference_image_urls | No | Pictures of the people, outfits or places the result should keep. |
resolution | No | Output size. Bigger sizes look sharper and cost more. 480p · 540p · 768p · Default: 768p |
seed | No | Reuse the same number to get a repeatable result. ≥ 0 · ≤ 2147483647 |
video_url | Yes | The clip to work on. |
Pricing
Prices are in USD. The API uses a separate USD balance. Website credits cannot be used or transferred to it, and API funds cannot be transferred to website credits. When a task is accepted, we reserve the maximum quoted amount from your API balance. On completion, we charge for actual usage, never more than the hold, and release the remainder. If the task fails, the hold is released.
| Option | USD per second |
|---|---|
| resolution 480p | $0.03 |
| resolution 540p | $0.045 |
| resolution 768p | $0.066 |
Send the same request body to POST /jobs/quote to see the estimated and maximum USD charge before running. A quote does not run the task or charge your balance.
Use cases
Change the weather, light or time of day
Turn a bright afternoon into dusk, add rain, or swap harsh midday sun for soft morning light, as the terrace example does. Because the people and the motion come from your clip, you can try several moods on one take instead of shooting again.
Restyle or reset a finished clip
Ask for a different style, season or setting and keep the performance. Name what must stay, such as the face, the pose and the movement. So far we have tested weather and light; style and setting changes are what the endpoint is built for, but check a draft before you rely on it.
Change one element
Aim the prompt at a specific element, such as the colour of a coat or an object in the scene, and say that everything else stays. Run a draft first and look at the parts you did not ask to change.
Prompt guide & pro tips
Say what changes, then what stays
The terrace prompt is a good shape: "Turn the sunny afternoon into a stormy dusk: dark clouds over the sea, rain streaking across the frame, her hair and coat wet and wind-blown, distant lightning. Keep her face, pose and movement." The first sentence is the change; the second is the guard rail. Write it as a change to the clip you sent, not as a new scene.
Describe the environment, not the movement
In the test run the prompt mentioned wind-blown hair, and the result added a hand lifting to her hair. If the motion in your clip is what you care about, describe the weather and the light and leave the actions to the source.
Draft low, then deliver at 768p
Treat a draft as a rough check of the idea, not a preview of the final look. If an effect reads lighter than you wanted, make the prompt more specific about how strong it should be and run a second take.
Match the length to your source
Set duration_seconds to the length of your clip, up to 15 seconds, to edit all of it.
API quick start
Call the MiniMax H3 Edit API in three steps
- Create an API key. Sign in to PopNow and create a key in the API console. Keys start with sk-popnow-. API calls use your separate USD balance.
- Send the request. POST /jobs/createTask with model minimax/h3-edit/video-edit and an input object that follows the parameter table. The response returns a task ID straight away.
- Get the result. Poll GET /jobs/recordInfo with the task ID, or pass a callBackUrl and we post the result to it. Download the output links when the task succeeds.
# 1. Create the task. Keep the Idempotency-Key if you retry after a network error.
export IDEMPOTENCY_KEY="$(uuidgen)"
curl -X POST https://api.popnow.ai/api/v1/jobs/createTask \
-H "Authorization: Bearer $POPNOW_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: $IDEMPOTENCY_KEY" \
-d '{
"model": "minimax/h3-edit/video-edit",
"input": {
"prompt": "A slow aerial shot drifting over a misty pine forest at sunrise",
"video_url": "https://media.popnow.ai/effects/flower-bloom-1abd6efb.mp4",
"resolution": "768p",
"duration_seconds": 5
}
}'
# 2. Copy data.taskId from the response, then poll until state is succeeded, failed, canceled or expired.
TASK_ID="TASK_ID_FROM_RESPONSE"
curl "https://api.popnow.ai/api/v1/jobs/recordInfo?taskId=$TASK_ID" \
-H "Authorization: Bearer $POPNOW_API_KEY"Replace the prompt and any links with your own. The API tab has the response, polling and callback reference.
What to build with the MiniMax H3 Edit API
- A one-tap restyle feature in which the user picks rain, snow, night or film, and the app sends the matching sentence as prompt.
- A pre-production tool that re-lights a location clip for several times of day without another shoot.
- A background job that edits the first 15 seconds of an uploaded clip for a social channel, with generate_audio set to false so the original audio stays.
MiniMax H3 Edit parameter notes
- video_url is required: an MP4 up to 15 seconds long. A longer clip is refused before any charge, so trim on the client first.
- duration_seconds takes whole seconds from 3 to 15, defaulting to 5. Set it to your source length to edit the whole clip. The source seconds are billed too.
- resolution takes 480p, 540p or 768p (the default). 768p is the native canvas. The lower tiers cost less per second and finish sooner, but the same prompt can look different at each tier.
- generate_audio defaults to true and gives the edit a new soundtrack. Set it to false to keep the source clip's own audio. Sound does not change the price.
- aspect_ratio is optional. Leave it out to keep the source shape, or pass 16:9, 9:16, 1:1, 4:3, 3:4, 21:9 or 9:21.
- reference_image_urls takes up to nine images, named <Picture 1> to <Picture 9> in the prompt by their list position. reference_audio_urls takes up to three clips, named <Audio 1> to <Audio 3>. Each reference adds a small per-reference charge.
- prompt is required, 3 to 5,000 characters. Write the change first and then what must stay, such as the face, the pose and the motion.
How MiniMax H3 Edit API pricing adds up
H3 Edit is billed in USD per second at the resolution rate: $0.030 per second at 480p, $0.045 at 540p and $0.066 at 768p. The bill covers the edited seconds plus the source seconds, rounded up to whole seconds, at the same rate, plus a small charge for each reference image or audio clip. A five-second source edited to five seconds at 768p therefore bills ten seconds, about $0.66 before references. Failed requests are not charged.
- resolution 480p, 5-second clip: $0.15 each, $15.00 for 100 clips
- resolution 768p, 5-second clip: $0.33 each, $33.00 for 100 clips
Async jobs, webhooks and rate limits
Every MiniMax H3 Edit request is asynchronous: createTask returns a taskId straight away, and the result arrives when the job finishes. Poll GET /jobs/recordInfo every few seconds, or pass a public https callBackUrl and verify the X-Webhook-Signature header (HMAC-SHA256 over taskId, timestamp and the body hash) before trusting the payload. Each account gets 1,000 write requests per 10 seconds; a 429 carries Retry-After. Send an Idempotency-Key with createTask so a retried request for minimax/h3-edit/video-edit never runs or bills twice.
Common MiniMax H3 Edit API errors
| errorCode | What it usually means and how to fix it |
|---|---|
invalid_request | duration_seconds outside 3 to 15, a resolution other than 480p, 540p or 768p, or more than nine reference images or three reference audio clips. |
invalid_asset | video_url could not be fetched, is not an MP4, or runs longer than 15 seconds. Trim the clip and upload it before calling the endpoint. |
content_rejected | The prompt or a reference was declined. Show the message and invite a rewording; the held balance is released automatically. |
unsupported_combination | A parameter pairing the endpoint does not accept. Drop the optional fields and retry with the defaults to find the field that causes the problem. |
MiniMax H3 Edit API questions
How much does the MiniMax H3 Edit API cost?
MiniMax H3 Edit costs $0.15 for a 5-second clip. Prices by option: 480p $0.03 per second, 540p $0.045 per second, 768p $0.066 per second. The API uses a separate USD balance. Website credits cannot be used or transferred to it, and API funds cannot be transferred to website credits. When a task is accepted, we reserve the maximum quoted amount from your API balance. On completion, we charge for actual usage, never more than the hold, and release the remainder. If the task fails, the hold is released.
What is the model name for MiniMax H3 Edit in the API?
Put minimax/h3-edit/video-edit for Edit Video in the model field of POST /jobs/createTask.
Is the source clip billed as well as the edit?
Yes. Source seconds are billed alongside the edited seconds, both at the rate for the resolution you pick. Set duration_seconds to the source length to edit the whole clip.
Can the edit keep the original soundtrack?
Yes. Set generate_audio to false and the clip returns with its original audio. The price is the same either way.
Why does 480p look different from 768p?
The lower tiers render on a smaller canvas. In one storm test the 480p trial showed heavier rain streaks than 768p did, so judge the look at the resolution you publish.
Can I edit a clip longer than 15 seconds?
No. Longer clips are refused before anything is charged. Trim the clip to 15 seconds or less and send the trimmed file.
What are the limits of the MiniMax H3 Edit API?
Clip length: 3–15 s; Resolutions: 480p · 540p · 768p; Aspect ratios: 16:9 · 9:16 · 1:1 · 4:3 · 3:4 · 21:9 · 9:21; Reference audio: 3; Reference images: 9. Requests outside these limits are rejected before anything is charged.
Can I try MiniMax H3 Edit without writing code?
Run MiniMax H3 Edit in the Playground with your API USD balance, without an API key. Each run appears in your developer history with its matching code.
All MiniMax models
Kling O3 also edits clips and works up to 4K. H3 Edit is the quicker, lower-priced choice for restyling and weather changes.
- MiniMax H3 APIExpressive clips with audio, up to 30 seconds or 2KFrom $0.125 for a 5-second clip
- MiniMax Speech 2.8 HD APIHigh-fidelity speech in 400+ voices with emotion controlFrom $0.12 per 1,000 characters
- MiniMax Music 3.0 APIFinished stereo songs, with vocals or instrumentalFrom $0.18 per run
- MiniMax H3 Singularity LoRA APICommunity-tuned H3 video that loads your own LoRAsFrom $0.30 for a 5-second clip
- MiniMax H3 Max APIFast clips with ambience, effects and musicFrom $0.25 for a 5-second clip
- MiniMax H3 LoRA APIMiniMax H3 video with sound and up to three of your LoRAsFrom $0.25 for a 5-second clip
- MiniMax H3 Image LoRA APIH3 stills and edits with up to three of your LoRAsFrom $0.042 per image
- MiniMax H3 Edit APIChange the light, weather or style of a clip; people and motion stayFrom $0.15 for a 5-second clip
- MiniMax Speech 2.8 Turbo APIFast speech in 400+ voices with emotion controlFrom $0.072 per 1,000 characters
- MiniMax H3 360 Orbit APIA full 360° camera orbit around the subject of one photoFrom $0.25 for a 5-second clip