unsloth / minimax family / text-generation
minimax-m2
minimax/minimax-m2
Run locally Smallest indexed GGUF / needs Ollama
ollama run hf.co/minimax/minimax-m2/MiniMax-M2-UD-TQ1_0.gguf Benchmarks
Where it scores.
General
69.9
Coding
56.5
Math
56.9
Reasoning
52.5
Normalized 0–100 across indexed models.
Runnable variants
Pick a size.
| Precision | Format | Size | Params | File |
|---|---|---|---|---|
| Q4 / smallest | GGUF | 56.4 GB | — | MiniMax-M2-UD-TQ1_0.gguf |
| IQ1_S | GGUF | 64.0 GB | — | UD-IQ1_S/MiniMax-M2-UD-IQ1_S-00001-of-00002.gguf |
| IQ1_M | GGUF | 68.7 GB | — | UD-IQ1_M/MiniMax-M2-UD-IQ1_M-00001-of-00002.gguf |
| IQ2_XXS | GGUF | 74.0 GB | — | UD-IQ2_XXS/MiniMax-M2-UD-IQ2_XXS-00001-of-00002.gguf |
| IQ2_M | GGUF | 78.1 GB | — | UD-IQ2_M/MiniMax-M2-UD-IQ2_M-00001-of-00002.gguf |
| Q2_K | GGUF | 83.3 GB | — | Q2_K/MiniMax-M2-Q2_K-00001-of-00002.gguf |
| Q2_K_L | GGUF | 83.4 GB | — | Q2_K_L/MiniMax-M2-Q2_K_L-00001-of-00002.gguf |
| Q2_K_XL | GGUF | 85.8 GB | — | UD-Q2_K_XL/MiniMax-M2-UD-Q2_K_XL-00001-of-00002.gguf |
| IQ3_XXS | GGUF | 93.6 GB | — | UD-IQ3_XXS/MiniMax-M2-UD-IQ3_XXS-00001-of-00002.gguf |
| Q3_K_S | GGUF | 98.7 GB | — | Q3_K_S/MiniMax-M2-Q3_K_S-00001-of-00002.gguf |
| Q3_K_XL | GGUF | 101.5 GB | — | UD-Q3_K_XL/MiniMax-M2-UD-Q3_K_XL-00001-of-00003.gguf |
| Q3_K_M | GGUF | 109.3 GB | — | Q3_K_M/MiniMax-M2-Q3_K_M-00001-of-00003.gguf |
| IQ4_XS | GGUF | 121.9 GB | — | IQ4_XS/MiniMax-M2-IQ4_XS-00001-of-00003.gguf |
| IQ4_NL | GGUF | 129.0 GB | — | IQ4_NL/MiniMax-M2-IQ4_NL-00001-of-00003.gguf |
| Q4_0 | GGUF | 129.5 GB | — | Q4_0/MiniMax-M2-Q4_0-00001-of-00003.gguf |
| Q4_K_S | GGUF | 130.0 GB | — | Q4_K_S/MiniMax-M2-Q4_K_S-00001-of-00003.gguf |
| Q4_K_XL | GGUF | 131.6 GB | — | UD-Q4_K_XL/MiniMax-M2-UD-Q4_K_XL-00001-of-00003.gguf |
| Q4_K_M | GGUF | 138.3 GB | — | Q4_K_M/MiniMax-M2-Q4_K_M-00001-of-00003.gguf |
| Q4_1 | GGUF | 143.2 GB | — | Q4_1/MiniMax-M2-Q4_1-00001-of-00003.gguf |
| Q5_K_S | GGUF | 157.5 GB | — | Q5_K_S/MiniMax-M2-Q5_K_S-00001-of-00004.gguf |
| Q5_K_XL | GGUF | 162.1 GB | — | UD-Q5_K_XL/MiniMax-M2-UD-Q5_K_XL-00001-of-00004.gguf |
| Q5_K_M | GGUF | 162.3 GB | — | Q5_K_M/MiniMax-M2-Q5_K_M-00001-of-00004.gguf |
| Q6_K | GGUF | 187.8 GB | — | Q6_K/MiniMax-M2-Q6_K-00001-of-00004.gguf |
| Q6_K_XL | GGUF | 194.3 GB | — | UD-Q6_K_XL/MiniMax-M2-UD-Q6_K_XL-00001-of-00004.gguf |
| Q8_0 | GGUF | 243.1 GB | — | Q8_0/MiniMax-M2-Q8_0-00001-of-00005.gguf |
| Q8_K_XL | GGUF | 261.4 GB | — | UD-Q8_K_XL/MiniMax-M2-UD-Q8_K_XL-00001-of-00006.gguf |
| BF16 | GGUF | 457.5 GB | — | BF16/MiniMax-M2-BF16-00001-of-00010.gguf |
Sizes are file sizes, not peak runtime memory.
Raw benchmarks
The inputs.
| Source | Task | Raw | Normalized |
|---|---|---|---|
| Arena | ArenaScore | 1343.5 | 80.4 |
| ZeroEval | AIME2025 | 78.0 | 78.0 |
| ZeroEval | GPQA | 78.0 | 78.0 |
| ZeroEval | HLE | 12.5 | 12.5 |
| ZeroEval | SWEBenchVerified | 69.4 | 69.4 |
| ZeroEval | SciCode | 36.0 | 36.0 |
From the Hugging Face model card
About this model.
From the Hugging Face model card