TPSTokens Per Second

Mistral-Small-4-119B-2603: hardware requirements

Mistral-Small-4-119B-2603 needs about 79 GB of memory at Q4 quantization — a 74 GB download plus room for context and overhead. That fits a machine with 128GB RAM.

Parameters
119B
Tier
Needs a workstation
Licence
See model card
Released
2026-05-28

Will it run on your machine?

8GB RAMNot enough memory16GB RAMNot enough memory32GB RAMNot enough memory64GB RAMNot enough memory128GB RAMRuns well8GB VRAMNot enough memory12GB VRAMNot enough memory16GB VRAMNot enough memory24GB VRAMNot enough memory32GB VRAMNot enough memory

Download sizes by quantization

Lower quantization means a smaller file and less memory, at some cost to quality. Q4_K_M is the usual starting point. What is this?

QuantDownloadRAM @ 4KRAM @ 32KSource
Q4_K_Mrecommended74 GB79 GB96 GBmeasured
Q5_K_M89 GB95 GB112 GBmeasured
Q6_K99 GB105 GB122 GBmeasured
Q8_0126 GB132 GB149 GBmeasured
BF16239 GB244 GB261 GBmeasured

“Measured” sizes are read from published GGUF files (unsloth/Mistral-Small-4-119B-2603-GGUF). “Estimated” sizes are derived from the parameter count and are typically within a few percent. The RAM columns add the context cache and runtime overhead to the download size.

Model card on HuggingFace →API specs & pricing →Verified 2026-08-01