How much VRAM does Mistral Small 4 119B need?

Mistral Small 4 119B (119B parameters) requires approximately 82.4 GB of memory with Q4_K_M quantization.

What is the best quantization for Mistral Small 4 119B?

The recommended quantization for Mistral Small 4 119B is Q4_K_M, which balances quality and memory efficiency.

Can it run?

No, Mistral Small 4 119B requires more memory than NVIDIA A16 64GB provides.

FWon't run

Too heavy

Using Q4_K_M in vLLM

Capabilities:

Fit status

Too heavy

Decode

17.3 tok/s

TTFT

11181 ms

Safe context

12K

Memory

82.4 GB / 64.0 GB

Weights72.6 GB

KV Cache1.0 GB

Runtime2.4 GB

Headroom6.4 GB

Workload	Grade	Fit	Decode	TTFT	Context
Agentic Coding	F	Too heavy	18.8 tok/s	15012 ms	25K
Chat	F	Too heavy	18.8 tok/s	5630 ms	6K
Coding	F	Too heavy	17.3 tok/s	11181 ms	12K
RAG	F	Too heavy	18.8 tok/s	18765 ms	25K
Reasoning	F	Too heavy	18.8 tok/s	12197 ms	12K

How Mistral Small 4 119B (119B params) fits at each quantization level on NVIDIA A16 64GB (64.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_KBest for your GPU	2	46.4 GB	Low	C45
Q3_K_S	3	58.3 GB	Low	C45

Upgrade options

MacBook Pro M3 Max 128GBBest value

C9.6 tok/s decode

~$2,499 MSRP

AMD Instinct MI350X 288GBBudget pick

C234.1 tok/s decode

~$8,000 MSRP

NVIDIA H200 141GBNVIDIA upgrade