mistralai/devstral-small-2-2512

mistral3 · 24B · 4bit

MacBook Air (Apple M4)

32 GB · macOS 26.3

Tested on March 4, 2026 · Submitted by Topaz750

Top 48% Compare

Global Score

69 /100

Good

Hardware Fit

50/100

Quality

77/100

Get this model

Get it in LM Studio

Search and download models directly from the app

🤗

Find on HuggingFace

MLX versions & conversions

Hardware

Machine: MacBook Air
CPU: Apple M4
Cores: 10 total (4 perf + 6 eff)
Frequency: 2.4 GHz
RAM: 32 GB LPDDR5
GPU: Apple M4
OS: macOS 26.3
Arch: arm64
Power Mode: balanced

Performance

Tokens/sec Tokens generated per second — higher is better: 5.7
Standard deviation Variation of speed across benchmark runs — lower means more consistent: ±3.1
First chunk latency Delay before the first streamed chunk arrives from runtime: 5.0 s
Time to first token How fast the model starts responding — lower is better: 5.0 s
Load time Time to load the model into memory before inference: N/A
Memory usage RAM consumed during inference — percentage of total system memory: 13.2 GB (41%)
Total tokens Total number of tokens generated across all benchmark prompts: 2820

Score breakdown

Speed

11/50

Time to first token

9/20

Memory

30/30

Quality

Reasoning

16/20

Coding

18/20

Instruction following

13/20

Structured output

14/15

Math

6/15

Multilingual

10/10

Category levels

Reasoning: Strong Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Weak Multilingual: Strong

Metadata

Spec version: 0.2.0
Runtime: LM Studio 0.4.6+1
Model format: MLX
Hardware profile: BALANCED
Result hash: 59e59e20e88c7210155b9a9b693dc13145fd8e704ecfe901b547a7c449ee74b8

Interpretation

Hardware fit: 50/100. Overall suitability: GOOD (Global 69/100). Category profile: Reasoning: Strong, Coding: Strong, Instruction Following: Adequate, Structured Output: Strong, Math: Weak, Multilingual: Strong.

Warnings

Token speed is unstable (stddev 3.1 tok/s, mean 5.7 tok/s) — may indicate thermal throttling or memory pressure.

Bench Environment

Power: AC CPU load: avg 11% (peak 16%)

Run yours now

$ npm install -g metrillm@latest

$ metrillm

Requires Node 20+ and Ollama or LM Studio running

Or run without installing: npx metrillm@latest