unsloth / mistral family
magistral-small-2506
mistral/magistral-small-2506
Run locally Smallest indexed GGUF / needs Ollama
ollama run hf.co/mistral/magistral-small-2506/Magistral-Small-2506-UD-IQ2_XXS.gguf Benchmarks
Where it scores.
General Provisional
65.9
Coding Provisional
65.5
Math Provisional
65.2
Reasoning Provisional
65.5
Normalized 0–100 across indexed models.
Runnable variants
Pick a size.
| Precision | Format | Size | Params | File |
|---|---|---|---|---|
| IQ2_XXS / smallest | GGUF | 6.8 GB | — | Magistral-Small-2506-UD-IQ2_XXS.gguf |
| IQ2_M | GGUF | 8.2 GB | — | Magistral-Small-2506-UD-IQ2_M.gguf |
| Q2_K | GGUF | 8.9 GB | — | Magistral-Small-2506-Q2_K.gguf |
| Q2_K_L | GGUF | 9.0 GB | — | Magistral-Small-2506-Q2_K_L.gguf |
| Q2_K_XL | GGUF | 9.3 GB | — | Magistral-Small-2506-UD-Q2_K_XL.gguf |
| IQ3_XXS | GGUF | 9.4 GB | — | Magistral-Small-2506-UD-IQ3_XXS.gguf |
| Q3_K_S | GGUF | 10.4 GB | — | Magistral-Small-2506-Q3_K_S.gguf |
| Q3_K_M | GGUF | 11.5 GB | — | Magistral-Small-2506-Q3_K_M.gguf |
| Q3_K_XL | GGUF | 11.9 GB | — | Magistral-Small-2506-UD-Q3_K_XL.gguf |
| IQ4_XS | GGUF | 12.8 GB | — | Magistral-Small-2506-IQ4_XS.gguf |
| IQ4_NL | GGUF | 13.5 GB | — | Magistral-Small-2506-IQ4_NL.gguf |
| Q4_0 | GGUF | 13.5 GB | — | Magistral-Small-2506-Q4_0.gguf |
| Q4_K_S | GGUF | 13.5 GB | — | Magistral-Small-2506-Q4_K_S.gguf |
| Q4_K_M | GGUF | 14.3 GB | — | Magistral-Small-2506-Q4_K_M.gguf |
| Q4_K_XL | GGUF | 14.5 GB | — | Magistral-Small-2506-UD-Q4_K_XL.gguf |
| Q4_1 | GGUF | 14.9 GB | — | Magistral-Small-2506-Q4_1.gguf |
| Q5_K_S | GGUF | 16.3 GB | — | Magistral-Small-2506-Q5_K_S.gguf |
| Q5_K_M | GGUF | 16.8 GB | — | Magistral-Small-2506-Q5_K_M.gguf |
| Q5_K_XL | GGUF | 16.8 GB | — | Magistral-Small-2506-UD-Q5_K_XL.gguf |
| Q6_K | GGUF | 19.3 GB | — | Magistral-Small-2506-Q6_K.gguf |
| Q6_K_XL | GGUF | 20.8 GB | — | Magistral-Small-2506-UD-Q6_K_XL.gguf |
| Q8_0 | GGUF | 25.1 GB | — | Magistral-Small-2506-Q8_0.gguf |
| Q8_K_XL | GGUF | 29.0 GB | — | Magistral-Small-2506-UD-Q8_K_XL.gguf |
| BF16 | GGUF | 47.2 GB | — | Magistral-Small-2506-BF16.gguf |
Sizes are file sizes, not peak runtime memory.
Raw benchmarks
The inputs.
| Source | Task | Raw | Normalized |
|---|---|---|---|
| ZeroEval | AIME2025 | 62.8 | 62.8 |
| ZeroEval | GPQA | 68.2 | 68.2 |
From the Hugging Face model card
About this model.
From the Hugging Face model card