DataLearner logo
MI

MiniMax H3

Multimodal model

MiniMax H3

Release date: 2026-07-31Updated: 2026-08-22 23:22:37.628413
Live demoGitHubHugging FaceCompare
Parameters
33B
Context length
15s
Chinese support
Supported
Reasoning ability

MiniMax H3 is a multimodal model from MiniMaxAI, released on 2026-07-31. It accepts text, image, audio, and video input and returns audio and video output. The cataloged parameter count is 33B. The recorded duration limit is 15s. Cataloged capabilities include Multilingual. The checkpoint is listed under the Minimax H3 Community License, with the license link included in the model references. Use the linked references to confirm current access, licensing, and provider-specific limits.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

MiniMax H3

Model basics

Reasoning traces
Not supported
Thinking modes
Thinking modes not supported
Context length
15s tokens
Max output length
No data
Model type
Multimodal model
Modality (in / out)
Text, Image, Audio, Video → Audio, Video
Release date
2026-07-31
Model file size
No data
MoE architecture
No
Total params / Active params
33B / N/A
Knowledge cutoff
No data
MiniMax H3

Open source & experience

Code license
Weights license
- Commercial use permitted
GitHub repo
https://github.com/MiniMax-AI/MiniMax-H3
Hugging Face
https://huggingface.co/MiniMaxAI/MiniMax-H3
Live demo
https://hailuoai.video/tools/minimax-h3
MiniMax H3

Official resources

Paper
Paper
DataLearnerAI blog
No blog post yet
MiniMax H3

API details

API speed
3/5
💡Default unit: $/1M tokens. If vendors use other units, follow their published pricing.
Learn about pricing modes
Standard
TypeConditionInputOutput
ImageRequest Type = more_than_5_input_images¥0.200/ img-
Audio-¥0.0000/ sec-
VideoRequest Type = resolution_768p¥0.500/ sec¥0.500/ sec
VideoRequest Type = resolution_2k¥0.800/ sec¥0.800/ sec
MiniMax H3

Benchmark Results

No benchmark data to show.

Compare with other models

No curated comparisons for this model yet.

Want a custom combination? Open the compare tool

MiniMax H3

Publisher

MiniMax H3

Model Overview

MiniMax H3 is a multimodal model from MiniMaxAI, released on 2026-07-31.

It accepts text, image, audio, and video input and produces audio and video output. Its cataloged capabilities include Multilingual. The cataloged parameter count is 33B. The recorded duration limit is 15s.

The checkpoint is listed under the Minimax H3 Community License, with the license link included in the model references. The page records 6 API pricing rules from MiniMaxAI; current provider pricing and conditions should be checked before deployment. The page links 4 release, model-card, repository, or provider references for checking the underlying claims. Specifications, availability, and prices can change; undisclosed values are intentionally left unstated.

MiniMax H3

FAQ

What is MiniMax H3?

MiniMax H3 is a multimodal model from MiniMaxAI, released on 2026-07-31. It accepts text, image, audio, and video input and returns audio and video output. The cataloged parameter count is 33B. The recorded duration limit is 15s. Cataloged capabilities include Multilingual. The checkpoint is listed under the Minimax H3 Community License, with the license link included in the model references. Use the linked references to confirm current access, licensing, and provider-specific limits.

What input and output modalities does MiniMax H3 support?

The current model record lists text, image, audio, and video as input and audio and video as output.

What are the main recorded specifications for MiniMax H3?

The cataloged parameter count is 33B. The recorded duration limit is 15s. Fields without a source-backed value remain undisclosed.

Does MiniMax H3 have API pricing?

The page records 6 API pricing rules from MiniMaxAI; current provider pricing and conditions should be checked before deployment.

Is MiniMax H3 open source?

The checkpoint is listed under the Minimax H3 Community License, with the license link included in the model references. Review the linked license text before commercial or derivative use.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.

DataLearner WeChat QR code