Mistral Small 3

January 2025 Mistral 24B. Llama 3.3 70B-tier quality at a third of the VRAM.

Parameters

24B

Family

Mistral

Context

32K tokens

FP16 weights

48GB

// where you can run it

// hugging face stats (cached daily)

76K downloads · 958 likes · license: apache-2.0 · updated 9 months ago

What you need to run this.

$ ./vrambudget --model mistral-small-3 --by quant

// budgets shown at ctx 8K, concurrency 1, 15% safety headroom. Tune in the calculator →

$ grep --params similar catalog.json

$ gh discussion list

// sign in with github to leave a comment. threads live in the repo's discussions tab.