How much VRAM does Devstral Small 1.1 need?

Devstral Small 1.1 (24B parameters) requires approximately 47.2 GB of memory with Q4_K_M quantization.

What is the best quantization for Devstral Small 1.1?

The recommended quantization for Devstral Small 1.1 is Q4_K_M, which balances quality and memory efficiency.

Can it run?

Yes, Mac Studio M3 Ultra 256GB can run Devstral Small 1.1 with a C grade (Runs well). Expected decode speed: 36.4 tok/s.

CUsable

Runs well

Using Q4_K_M in Ollama

Capabilities:

Fit status

Runs well

Decode

36.4 tok/s

TTFT

5324 ms

Safe context

62K

Memory

47.2 GB / 184.3 GB

Weights14.6 GB

KV Cache3.8 GB

Runtime1.2 GB

Headroom27.6 GB

Workload	Grade	Fit	Decode	TTFT	Context
Agentic Coding	C	Runs well	36.4 tok/s	7744 ms	116K
Chat	C	Runs well	36.4 tok/s	2904 ms	33K
Coding	C	Runs well	36.4 tok/s	5324 ms	62K
RAG	C	Runs well	36.4 tok/s	9680 ms	116K
Reasoning	C	Runs well	36.4 tok/s	6292 ms	62K

How Devstral Small 1.1 (24B params) fits at each quantization level on Mac Studio M3 Ultra 256GB (184.3 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	9.4 GB	Low	D30
Q3_K_S	3	11.8 GB	Low	D31
NVFP4	4

Ollama

ollama run devstral-small-2507

HuggingFace

huggingface-cli download devstral-small-2507

Upgrade options

AMD Instinct MI350X 288GBBudget pick

C398.9 tok/s decode

~$8,000 MSRP

AMD Instinct MI300X 192GBBest value

C282.4 tok/s decode

~$15,000 MSRP

H100 NVL 188GBBiggest leap