Notice on MiniMax-M3 Model Price Reduction
To continuously reduce user costs, CloudBase will reduce the price of the MiniMax-M3 model (model parameter: minimax-m3) starting from July 6, 2026, 16:00 (Beijing Time). Input, output, and cache hit fees will all be reduced by 50%, allowing you to enjoy the same high-quality model service at a lower price.
This adjustment only involves price changes; model service capabilities remain unchanged.
Adjustment Details
| Model | Tier (tokens) | Input (CNY/M tokens) | Output (CNY/M tokens) | Cache Hit (CNY/M tokens) |
|---|---|---|---|---|
| MiniMax-M3 | Input length (0, 512k] | From 4.2 to 2.1 | From 16.8 to 8.4 | From 0.84 to 0.42 |
| MiniMax-M3 | Input length 512k+ | From 8.4 to 4.2 | From 33.6 to 16.8 | From 1.68 to 0.84 |
Thank you for your trust and support! We will continue to optimize our model service pricing strategy to provide you with more cost-effective model services.
If you encounter any issues, please visit the CloudBase Community for assistance.