How much VRAM does Devstral Small 1.1 need?

Devstral Small 1.1 (24B parameters) requires approximately 26.5 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 mini M4 64GB can run Devstral Small 1.1 with a C grade (Runs well). Expected decode speed: 5.3 tok/s.

CUsable

Runs well

Using Q4_K_M in Ollama

Capabilities:

Fit status

Runs well

Decode

5.3 tok/s

TTFT

36337 ms

Safe context

28K

Memory

26.5 GB / 46.1 GB

Weights14.6 GB

KV Cache3.8 GB

Runtime1.2 GB

Headroom6.9 GB

Workload	Grade	Fit	Decode	TTFT	Context
Agentic Coding	C	Runs well	5.3 tok/s	52854 ms	49K
Chat	C	Runs well	5.3 tok/s	19820 ms	15K
Coding	C	Runs well	5.3 tok/s	36337 ms	28K
RAG	C	Runs well	5.3 tok/s	66068 ms	49K
Reasoning	C	Runs well	5.3 tok/s	42944 ms	28K

How Devstral Small 1.1 (24B params) fits at each quantization level on Mac mini M4 64GB (46.1 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	9.4 GB	Low	D34
Q3_K_S	3	11.8 GB	Low	D35
NVFP4	4

Ollama

ollama run devstral-small-2507

HuggingFace

huggingface-cli download devstral-small-2507

Upgrade options

MacBook Pro M4 Max 96GBBudget pick

C23.5 tok/s decode

~$2,499 MSRP

Mac Studio M2 Ultra 128GBBest value

C31.7 tok/s decode

~$3,999 MSRP

RTX PRO 5000 Blackwell 48GBBiggest leap