MiniMax API v1
September 25, 2024 (February 27, 2026)
This is experimental API for for the MiniMax AI, a Chinese AI startup backed by Alibaba and Tencent.
We provide full API support for following MiniMax/Hailuo AI models:
- MiniMax videos supports 15 models across 3 families: Hailuo 01/02/2.3, Veo 3.1 (including S2V subject reference), and Sora 2. These models are notable for generating ultra-realistic video clips with precise prompt following.
- MiniMax images supports 8 models: Midjourney V7, Midjourney NiJi 7, Nano Banana 2, Nano Banana Pro, GPT Image 1.5, Seedream 4.5/5.0, and image-01.
- MiniMax agent supports multiple AI models including video generation (Hailuo 2.3, 02, Veo 3.1, Sora 2), image generation (Midjourney V7/NiJi 7, Nano Banana 2/Pro, Seedream 4.5/5.0, Kontext, Qwen, Kolors), and audio synthesis (Speech 2.5, Music 2.0).
Nano Banana 2/Pro deployment provided by MiniMax has less content moderation compared to Google Flow and is capable of editing photos of minors or famous people.
π² Credits estimator
Actual costs may change, see hailuoai.video for current pricing. With the $79.99/m Master account (12,000 credits), you can generate ~12,000 songs or up to 12,000 images.
Images β per image, see images/create for details
| Model | Credits |
|---|---|
| image-01 | 1 |
| midjourney-v7, midjourney-niji7 | 3 |
| seedream-4.5, seedream-5.0 | 4 |
| nano-banana-2 (512P/1K) | 4 |
| nano-banana-2 (2K) | 5 |
| nano-banana-2 (4K) | 8 |
| nano-banana-pro (1K/2K) | 6 |
| nano-banana-pro (4K) | 10 |
| gpt-image-1.5 (Low/Medium/High) | 4 / 8 / 15 |
Videos β per video, see videos/create for details
| Model | Credits |
|---|---|
| Hailuo 01 (all models) | 25 |
| 02, T2V-2.3, I2V-2.3 (768p/1080p) | 25-80 |
| I2V-2.3-Fast (768p/1080p) | 15-50 |
| Sora-2 (4/8/12sec) | 40 / 80 / 120 |
| Veo-3.1, Veo-3.1-S2V (720p/1080p/4K) | 120 / 120 / 180 |
| Veo-3.1-Fast, Veo-3.1-S2V-Fast (720p/1080p/4K) | 60 / 60 / 90 |
Audio β 1 credit per song
Postman collection (January 20, 2026)
LLM-friendly API spec Feed this to your LLM to build integrations
Q&A: How is your experimental MiniMax API different from the official MiniMax API?
Blogs:
- 16+ AI Image Models: The Showdown
- Nano Banana 2
- MiniMax Sora 2 & Midjourney V7
- MiniMax Sora 2, Veo 3.1
- MiniMax Agent: Nano Banana 2/Pro, Speech 2.5, Music 2.0 and moreβ¦
- MiniMax 2.3
- MiniMax 02 End Frame
- Groundbreaking Video Agents
- MiniMax Images
Articles:
Developer Community:
Table of contents
-
GETaccounts -
GETaccounts/account -
POSTaccounts -
DELaccounts/account -
GETfeatures -
POSTvideos/create -
POSTvideos/agent-create -
GETvideos/videoId -
GETvideos -
GETvideos/agent-templates -
GETvideos/characters -
POSTvideos/cancel -
DELvideos/videoId -
POSTimages/create -
GETimages -
GETimages/imageId -
DELimages/imageId -
POSTagent -
GETagent/jobId -
GETagent/jobs -
POSTfiles -
GETfiles -
DELfiles/fileID -
GETscheduler -
GETscheduler/available -
DELscheduler/id