MiniMax M3 models
4 MiniMax M3 models on Hugging Face, from 3.1B to 440.3B parameters. At the precision each one is published in, the smallest needs about 6.9 GB of VRAM (MiniMax-M3-EAGLE3-GQA, cheapest live fit: RTX 5060 Ti) and the largest about 955 GB (MiniMax-M3). The cheapest way to run MiniMax-M3-EAGLE3-GQA is $0.110/hr.
MiniMax M3 models
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| MiniMax-M3-EAGLE3-GQA | 3.1B | BF16 | 6.9 GB | RTX 5060 Ti | $0.110/hr |
| MiniMax-M3-EAGLE3 | 3.3B | BF16 | 7.3 GB | RTX 5060 Ti | $0.110/hr |
| MiniMax-M3 | 427.0B | BF16 | 955 GB | No live fit | – |
| MiniMax-M3-MXFP8 | 440.3B | F8_E4M3 | 492 GB | RTX PRO 6000 × 6 | $8.25/hr |
VRAM is for the precision the model is published in, with the same overhead and an 8,192-token context assumed on every page; see the methodology. The fit shown is the lowest-priced single GPU type that holds the model at that precision, or the lowest-priced multi-GPU set (up to 8) when none does.
More from MiniMax
- MiniMax M2 (4 models)
