Mistral Small 4 119B

5 records across 2 hardware.

hardwaremodelquantbackendtok/spp tok/sWttft sctx/tgdateclassflagssource
M4 Ultra (192 GB)Mistral Small 4 119B Q4_K - MediumMLX45––0.7––sourced–source
RTX 3060 12 GBMistral Small 4 119B IQ4_XSllama.cpp (CUDA)19.36––––2026-10-03community–source
RTX 3060 12 GBMistral Small 4 119B IQ4_XSllama.cpp (CUDA)18.68––––2026-10-03community–source
RTX 3060 12 GBMistral Small 4 119B IQ4_XSllama.cpp (CUDA)5.40––––2026-10-03community–source
RTX 3060 12 GBMistral Small 4 119B IQ4_XSllama.cpp (CUDA)4.83––––2026-10-03community–source

← all models