← Devstral Small

Devstral Small 2507

No machine on this network has measured this model yet. Everything below is what the catalog knows exists — where the weights come from, how large each build is, and whose work it is. Nothing here is a claim about what the model can do; that only appears once a machine has run it and signed the result.

Be the first to measure it It needs about 6.2 GB; 15 of 18 kinds of machine on the network can hold it. The first machine to sign a result takes 10 points on the providers board.What is measured in Devstral Small The members of this family that machines have already run, with their speed, price and context.

What it fits on

The smallest build needs about 6.2 GB; 15 of 18 kinds of machine on the network can hold it. Bars: each kind's memory (its card's, else RAM) · ticks: every build, at the size it held plus a fifth for the prompt.

  1. Linux, CPU only · 2 CPUs1 GB RAMnone
  2. NVIDIA GeForce RTX 3050 Laptop GPU4 GB on the cardnone
  3. NVIDIA GeForce GTX 16504 GB on the cardnone
  4. AMD Radeon(TM) Vega 8 Mobile Graphics7 GB RAM1 of 12
  5. Windows, CPU only · 12 CPUs7 GB RAM1 of 12
  6. Apple silicon8 GB RAM2 of 12
  7. Apple M18 GB RAM2 of 12
  8. Mac, CPU only · 8 CPUs8 GB RAM2 of 12
  9. NVIDIA GeForce RTX 3080 Ti12 GB on the card3 of 12
  10. Tesla T415 GB on the card4 of 12
  11. Intel(R) UHD Graphics 73016 GB RAM4 of 12
  12. Intel(R) Iris(R) Xe Graphics16 GB RAM4 of 12
  13. Apple M416 GB RAM4 of 12
  14. Apple M116 GB RAM4 of 12
  15. NVIDIA L422 GB on the card7 of 12
  16. 2× Tesla T429 GB on the card10 of 12
  17. Linux, CPU only · 4 CPUs31 GB RAM10 of 12
  18. NVIDIA RTX PRO 6000 Blackwell Server Edition95 GB on the card11 of 12
hf.co/unsloth/Devstral-Small-2507-GGUF:IQ1_Sollama · IQ1_S · 5.2 GiB · quantizer (unsloth)
hf.co/unsloth/Devstral-Small-2507-GGUF:IQ2_XXSollama · IQ2_XXS · 6.3 GiB · quantizer (unsloth)
hf.co/unsloth/Devstral-Small-2507-GGUF:IQ3_XXSollama · IQ3_XXS · 8.8 GiB · quantizer (unsloth)
lmstudio-community/Devstral-Small-2507-MLX-4bitmlx · 4BIT · 12.3 GiB · quantizer (lmstudio-community)
hf.co/mistralai/Devstral-Small-2507_gguf:Q4_K_Mollama · Q4_K_M · 13.3 GiB · vendor (mistralai)
hf.co/mistralai/Devstral-Small-2507_gguf:Q5_K_Mollama · Q5_K_M · 15.6 GiB · vendor (mistralai)
lmstudio-community/Devstral-Small-2507-MLX-6bitmlx · 6BIT · 17.8 GiB · quantizer (lmstudio-community)
RedHatAI/Devstral-Small-2507-quantized.w8a8vllm · W8A8 · 23.2 GiB · quantizer (RedHatAI)
lmstudio-community/Devstral-Small-2507-MLX-8bitmlx · 8BIT · 23.3 GiB · quantizer (lmstudio-community)
hf.co/mistralai/Devstral-Small-2507_gguf:Q8_0ollama · Q8_0 · 23.3 GiB · vendor (mistralai)
hf.co/mistralai/Devstral-Small-2507_gguf:BF16ollama · BF16 · 43.9 GiB · vendor (mistralai)
mistralai/Devstral-Small-2507vllm · BF16 · 87.8 GiB · vendor (mistralai)
derivatives — somebody else's change to these weights

Fine-tunes, merges and abliterations of this member, each its own model with its own page. None of their numbers are counted above.

Devstral-Small-2507-MLX-bf16not measuredDevstral-Small-2507-unsloth-bnb-4bitnot measured