MiniMax: MiniMax H3 Max

Hailuo H3-Max generates 5–15-second videos at 480p or 768p from text, a first frame, or first and last frames. Its speed-focused workflow is useful for quickly turning a shot idea into a short clip.

Compare models

hailuo-h3-max

AvailableMiniMaxVideoAsync

Price: From $0.05
Modalities: Video

About MiniMax H3 Max

MiniMax H3 Max is the speed-focused variant of MiniMax's H3 video model, co-developed by MiniMax with a partner and tuned for fast generation. It makes short clips from text, a first frame, or a first and last frame, at 480p or 768p. Compared with the standard H3, it gives up the 2K ceiling in exchange for quicker turnaround, so it fits drafting and idea testing.

Where it works well

It is tuned for generation speed, so shot ideas can be tested in quick succession.
First-frame and first-plus-last-frame inputs let you control where a clip starts and ends.
Whole-second durations from five to fifteen seconds are offered, giving more length range than Hailuo 2.3.
Text-to-video, image-to-video and start-end-to-video all work on one model ID.
The 480p setting gives a cheap draft path before committing to a higher-resolution render.

When to choose another model

Resolution stops at 768p; the standard MiniMax H3 is the one that reaches 2K.
It takes text, a first frame, or first and last frames only; reference-to-video is not available on this model.
It is an optimized draft tool, so final hero shots are better rendered on the standard H3.

Getting started

Create API key
Create a key in Console, then use it with every model on the platform.

Send your first request

Copy the example for your language and run it against the endpoint.

Image to videoStart and end frames to videoText to videoTokenLab endpoint

POST/v1/videos/generations

curl -X POST "https://api.tokenlab.sh/v1/videos/generations" \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
  "aspect_ratio": "16:9",
  "duration": 5,
  "operation": "text-to-video",
  "resolution": "768p",
  "model": "hailuo-h3-max",
  "prompt": "Cinematic sunrise over a calm lake with gentle camera motion."
}'

Parameters

Image to video

Parameter	Values and limits	Default	Notes
`aspect_ratio`	16:9, adaptive	Not provided	Not provided
`duration`	5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15	Not provided	Not provided
`image`	Not provided	Not provided	The current public contract does not specify an input count limit for this field. Do not infer a count from the example request.
`image_url`	Not provided	Not provided	The current public contract does not specify an input count limit for this field. Do not infer a count from the example request.
`model`	Not provided	Not provided	Not provided
`operation`Required	Not provided	Not provided	Not provided
`prompt`	Not provided	Not provided	Not provided
`resolution`	480p, 768p	Not provided	2K output is outside this public contract. The omitted-field default is not verified.
`start_image`	Not provided	Not provided	The current public contract does not specify an input count limit for this field. Do not infer a count from the example request.

A non-empty prompt is required for every operation, including first/last-frame inputs.

Start and end frames to video

Parameter	Values and limits	Default	Notes
`aspect_ratio`	16:9, adaptive	Not provided	Not provided
`duration`	5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15	Not provided	Not provided
`end_image`	Not provided	Not provided	The current public contract does not specify an input count limit for this field. Do not infer a count from the example request.
`image`	Not provided	Not provided	The current public contract does not specify an input count limit for this field. Do not infer a count from the example request.
`image_url`	Not provided	Not provided	The current public contract does not specify an input count limit for this field. Do not infer a count from the example request.
`model`	Not provided	Not provided	Not provided
`operation`Required	Not provided	Not provided	Not provided
`prompt`	Not provided	Not provided	Not provided
`resolution`	480p, 768p	Not provided	2K output is outside this public contract. The omitted-field default is not verified.
`start_image`	Not provided	Not provided	The current public contract does not specify an input count limit for this field. Do not infer a count from the example request.

A non-empty prompt is required for every operation, including first/last-frame inputs.

Text to video

Parameter	Values and limits	Default	Notes
`aspect_ratio`	16:9	Not provided	Not provided
`duration`	5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15	Not provided	Not provided
`model`	Not provided	Not provided	Not provided
`operation`Required	Not provided	Not provided	Not provided
`prompt`	Not provided	Not provided	Not provided
`resolution`	480p, 768p	Not provided	2K output is outside this public contract. The omitted-field default is not verified.

A non-empty prompt is required for every operation, including first/last-frame inputs.

No 2K or multimodal reference mode. Native audio has no separate public audio-disable selector.

Pricing

The Verified price applies to Verified, which costs less on most models. The Official price is the model maker's published price and applies to the more reliable Official route. Auto bills the route that completes the request.

Text to video / Image to video / Start and end frames to video · 480p

per second

Official price: $0.05
Verified price: —
Discount: —

Text to video / Image to video / Start and end frames to video · 768p

per second

Official price: $0.08
Verified price: —
Discount: —

	Official priceper second	Verified priceper second	Discount
Text to video / Image to video / Start and end frames to video · 480p	$0.05	—	—
Text to video / Image to video / Start and end frames to video · 768p	$0.08	—	—

Usage & activity

Success rate is the share of requests that completed. Latency is how long a full response takes; P95 means 95% of requests finished within that time.

Usage & availability

Last 24 hours

Requests
Success rate
P95 latency
Total tokens

Model performance

Last updated Nov 15, 2026, 1:48 AM UTC

Metrics appear once privacy and data volume thresholds are met.

Data is based on aggregate user requests, excluding status checks.

Open in Console

Open MiniMax H3 Max in Console with a prompt ready to edit or send.

Help me create with hailuo-h3-max using text-to-video at /v1/videos/generations. Show the result, status, and cost.

Use cases

Best for

Text to Video

Rapid shot drafting
Try five phrasings of one scene at 480p, then keep the clip whose motion reads best.
Frame-to-frame transitions
Set a first and last frame and let the model fill in the movement between two compositions.
Animating key art
Start from a poster or product image and add a short camera move.
Pre-visualization
Block out sequence timing with long clips before investing in full-resolution renders.

Prompt examples

From the photo of the empty beach, waves roll in and the sun sets, camera slowly pans left.

First frame: the city at dawn. Last frame: the same city at night. Time-lapse feel, smooth transition.

A cat jumps onto a kitchen counter and knocks over a glass, handheld camera, 12 seconds.

FAQ

What is the difference between H3 Max and H3?

H3 Max is the speed-optimized variant with 480p and 768p output and no 2K option. The standard H3 offers the higher resolution range and also accepts reference media, which Max does not.

Which inputs does it take?

Text, a first frame, or a first and last frame, for text-to-video, image-to-video and start-end-to-video. Reference media is not accepted by this variant. Reference media is only accepted by the standard H3.

How long can clips be?

MiniMax documents five to fifteen seconds in whole-second steps for H3 Max. Whole-second steps are available across that range. The range is wide enough for both quick tests and longer previews.

Who developed H3 Max, and for what goal?

MiniMax's documentation says the model was developed jointly with an outside partner and optimized for high-speed generation, at the cost of a lower maximum resolution than the standard H3.

When should I pick H3 Max?

Pick it for drafts, prompt testing and long low-resolution clips. Move to the standard H3 when the final output needs 2K or reference media inputs.

How much does MiniMax H3 Max cost?

On TokenLab, MiniMax H3 Max costs $0.05-$0.08 per second. The pricing table above shows the full breakdown. Rates depend on the billing unit, specification, and usage. Compare matching conditions in the model's detailed pricing; a single rate does not determine the total cost.

Which endpoint should MiniMax H3 Max use?

Use https://api.tokenlab.sh/v1/videos/generations for MiniMax H3 Max. The request example below shows the matching code shape.

Which operations does MiniMax H3 Max support?

MiniMax H3 Max supports Image to video, Start and end frames to video, Text to video. Select an operation above to see its endpoint and request example.