MiniMax · Video model
H3.
MiniMax H3 is a lightweight, open-weights video generation model from MiniMax. It is designed for precise multimodal editing and controlled content generation, including instruction-guided edits, text and brand rendering, and...
Balanced quality
—
Limited evidence
Text to video
—
No comparable result
Image to video
—
No comparable result
Price
—
Per minute at the evaluated setting
What this means.
This model has evidence for only one video workflow, so its recommendation confidence is reduced rather than assuming the missing result.
Practical specifications.
- Resolutions
- 2K
- Aspect ratios
- 21:9, 16:9, 4:3, 1:1, 3:4, 9:16
- Durations
- 5–15 seconds
- Generates audio
- Yes
- First/last frame control
- first_frame, last_frame
See where H3 stands.
| Model | Company | Quality | Evidence |
|---|---|---|---|
| Seedance 2.0 | ByteDance | 98 | Both workflows / exact match |
| HappyHorse 1.1 | Alibaba | 88 | Both workflows / exact match |
| HappyHorse 1.0 | Alibaba | 82 | Both workflows / exact match |
| Grok Imagine Video 1.5 | xAI | 80 | Limited |
| Wan 2.7 | Alibaba | 78 | Both workflows / exact match |
| Video v3.0 Pro | Kling AI | 70 | Both workflows / exact match |
| Veo 3.1 | 68 | Both workflows / exact match | |
| Video v3.0 Standard | Kling AI | 60 | Both workflows / exact match |
| Veo 3.1 Fast | 58 | Both workflows / exact match | |
| Grok Imagine Video | xAI | 52 | Both workflows / exact match |
| Veo 3.1 Lite | 46 | Both workflows / exact match | |
| Seedance 1.5 Pro | ByteDance | 28 | Both workflows / exact match |