Are MiniMax H3 and Hailuo 03 the same model?

Yes. Both names refer to the same release. Coverage uses either name; on reAPI the model id is minimax-h3.

Is MiniMax H3 free or paid?

MiniMax H3 is a paid model billed per second of generated video, with reference-video seconds added and images beyond the fifth billed per image. Eligible new accounts may receive promotional test credits; availability and amount are shown at signup or in the dashboard. See the pricing table above for current rates.

Does MiniMax H3 require a monthly subscription?

No. reAPI is pay-as-you-go: $1 buys 1,000 credits, credits never expire, and you only pay for completed generations. Failed MiniMax H3 tasks are refunded automatically.

How do I get a MiniMax H3 API key?

Create a reAPI account, open Settings → API Keys, and generate a key. The same key works for MiniMax H3 and every other model on the platform.

What inputs does MiniMax H3 reference-to-video accept?

Up to nine images, three videos, and three audio files per request. Videos are MP4/MOV, 2–15 seconds each with all clips totaling 15 seconds or less; audio is WAV/MP3 and cannot be used without an image or video reference.

Does MiniMax H3 generate audio?

Yes — MiniMax H3 renders native stereo audio together with the visuals. Direct dialogue, music, ambience, and sound effects in the same prompt that describes the scene.

How long can a MiniMax H3 video be?

4 to 15 seconds, in whole-second increments; the default is 6 seconds. Reference videos add their probed duration to the bill but not to the output length.

Which aspect ratios does MiniMax H3 support?

21:9, 16:9, 4:3, 1:1, 3:4, and 9:16. Text-to-video requires an explicit ratio; reference mode defaults to adaptive; image mode derives framing from your frames.

How does MiniMax H3 pricing work?

Per second of generated video at a flat 2K rate, plus the probed duration of any reference videos and a per-image charge for reference images beyond the first five. Audio input is free. See the table above for live rates.

How do I integrate MiniMax H3 into my app?

POST /api/v1/videos/generations with model minimax-h3 and your inputs, then poll /api/v1/tasks/:id until the status succeeds and output.video_urls arrives. Copy-paste examples in four languages live in the API docs.

MiniMax H3 — Multimodal Video Generation with Native Audio

MiniMax H3, also released as Hailuo 03, turns text, images, video clips, and audio references into 2K video with native stereo sound. Generate 4–15 second clips from a prompt, animate first and last frames, or direct up to fifteen reference files in a single MiniMax H3 request.

Input

Prompt*

Create a 15-second cinematic contemporary-dance scene in a quiet historic courtyard shortly after rainfall. The courtyard has pale stone arches, weathered columns, shallow puddles, and soft reflections across the wet ground. A young woman stands alone in the center wearing a flowing dark red midi dress with long sleeves and simple black dance shoes. Preserve her face, hairstyle, clothing, and body proportions throughout the entire performance.
[0–3 seconds]
Begin in a restrained medium-wide shot from behind. The dancer stands completely still while water drips from the surrounding arches. A soft piano note begins. She slowly turns her head, then rotates her shoulders and body toward the camera with controlled natural movement.
[3–7 seconds]
She begins a contemporary dance phrase with one extended arm, a grounded step forward, a low turn, and a smooth recovery into an upright pose. Her dress follows the movement with believable weight and delayed fabric motion. Keep both feet connected naturally to the wet stone surface without sliding.
[7–11 seconds]
The rhythm becomes stronger. Track laterally as she performs two connected turns, opens both arms, lowers briefly through the knees, and rises into a longer traveling movement across the courtyard. Her reflection should remain aligned with her position and motion in the puddles.
[11–15 seconds]
Move slightly closer as she completes one final controlled rotation and stops in a stable pose with one arm raised and the other lowered. The music cuts to a quiet sustained note. Hold the final posture while fabric and loose hair settle naturally and rain continues dripping in the background.
Camera: measured dolly movement, restrained lateral tracking, natural perspective, no rapid orbiting or abrupt zooms.
Lighting: soft overcast daylight with subtle warm reflections from the surrounding stone.
Audio: original minimal piano and low strings, light rainfall, fabric movement, footsteps on wet stone, and natural courtyard ambience.
Maintain realistic anatomy, continuous choreography, stable facial identity, accurate hand and foot placement, and physically believable clothing motion.
No extra dancers, cuts to unrelated locations, duplicated limbs, distorted hands, floating feet, sudden costume changes, exaggerated acrobatics, theatrical facial expressions, subtitles, text, or logos.

1 to 7,000 characters

Aspect ratio*

adaptive works in reference mode only

Duration (seconds)6

4 to 15 · default 6

Estimated cost1,095 credits≈ $1.095

Result

Try one of these prompts

What is the MiniMax H3 API?

MiniMax H3 is MiniMax's multimodal video model, also released as Hailuo 03 — both names refer to the same system. On reAPI it runs behind one async endpoint that routes by request shape: a prompt alone for text-to-video, first and last frames for image-to-video, or up to nine reference images, three reference videos and three audio tracks for reference-driven generation. Every request returns 4–15 seconds of video at up to 2K with native stereo audio, billed per second of output plus a small per-image charge for the sixth through ninth reference image.

What you can build with this model

Real-world workflows and production use cases you can build and ship with this model.

MiniMax H3 product film with consistent materials across shots

Product films with structured shots

MiniMax H3 builds commercials with varied camera angles, close-up feature shots, readable specifications, and consistent product materials across the whole sequence. Feed product photos in as references and MiniMax H3 keeps packaging, proportions, and brand identity stable from shot to shot.

Try MiniMax H3 now

MiniMax H3 cinematic scene with synchronized stereo audio

Cinematic scenes with directed sound

Direct dialogue, music, ambience, and sound effects in the same prompt as the visuals. MiniMax H3 generates native stereo audio together with the image stream, so lips, footsteps, and score stay synchronized without a separate audio pass.

MiniMax H3 reference-to-video combining images, video, and audio

Multimodal remix from references

Assign each asset a role: identity from an image, motion from a video clip, voice from an audio file. MiniMax H3 reference-to-video supports subject replacement, motion transfer, background modification, and relighting — one request, up to fifteen reference files.

Pricing

Credit-based — 1 credit = $0.001 USD. Pay only for completed generations.

First test cost

$0.05555 credits

One sample at the cheapest tier (1 image (6th-9th)).

Testing budget guide

Add credits

$10

≈ 181 tests

$50

≈ 909 tests

$100

≈ 1818 tests

Model	Category	Price
Standard	Video output (2K), per second	$0.183 183 credits · 1 second
Standard	Extra reference image (6th–9th)	$0.055 55 credits · 1 image (6th-9th)

Why reAPI

One model, three modes

Text-to-video, image-to-video, and reference-to-video share one endpoint and one MiniMax H3 model id. The request shape picks the mode — no separate integrations to maintain.

Native stereo audio

MiniMax H3 renders sound with the picture instead of bolting it on afterwards. Dialogue, ambience, music, and effects follow the same prompt that directs the camera.

Pay-as-you-go pricing

MiniMax H3 bills per second of output at one flat 2K rate. No subscription, no idle cost — $1 buys 1,000 credits that never expire, and failed renders refund automatically.

MiniMax H3 vs Veo 3.1

How MiniMax H3 on reAPI compares to Google's Veo 3.1 for prompt-driven video generation with native audio.

Capability

MiniMax H3 on reAPI

Veo 3.1

Clip length

4–15 seconds, whole-second steps

4, 6, or 8 seconds (remix extends to 15)

Max resolution

Up to 2K

Up to 4K on official channels

Native audio

Stereo audio always generated with the visuals, directed in the prompt

Optional synthesized audio (separate pricing tier)

Reference inputs

Up to 9 images + 3 videos + 3 audio files, each assignable to a role

Frame and image reference modes; no video or audio references

Editing controls

Subject replacement, motion transfer, background modification, relighting

Remix and frame-guided extension

Physical realism

Structured, commercial-grade shot progression

Top-tier physical realism and cinematic consistency

Comparison reflects publicly documented behavior at the time of writing. Capabilities and limits change; check each provider's current documentation before choosing.

Ship in three steps

step 01
Get an API key
Create a reAPI account and mint a key in the dashboard. The same key unlocks MiniMax H3 and every other model in the catalog.
Open
step 02
Test in the playground
Open the MiniMax H3 playground on this page, pick a mode, and submit a prompt or reference set without writing a line of code.
Open
step 03
Call the endpoint
POST to /api/v1/videos/generations with model minimax-h3, then poll the task id until the video URL lands.
Open

docs/api/minimax-h3

API reference

Drop-in code and the full parameter table.

View full reference

curl -X POST https://reapi.ai/api/v1/videos/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "minimax-h3",
    "prompt": "A cat walking slowly on the beach at sunset, cinematic shot",
    "aspect_ratio": "16:9",
    "duration": 6
  }'

Parameter validation failed

Rejected synchronously with the failing field named. Check enum values, ranges, and URL fields — nothing is charged.

Authentication or balance issue

401 means a missing or invalid Bearer key; 402 means the reserve exceeds your balance. Manage keys and credits in the console.

Content or material rejected

Reference material and outputs pass automated moderation. Rejected tasks fail with a clear error and a full refund.

Task failed or timed out

A task that reaches a failed state is never charged — the reserve refunds automatically. Keep the task id and retry when ready.

Frequently asked questions

Common questions about this model.

MiniMax H3 is MiniMax's general-purpose multimodal video model, also released as Hailuo 03. It generates 4–15 second videos at up to 2K resolution with native stereo audio from text, images, video clips, and audio references.