Back to leaderboard

qwen3.5:4b

OLLAMA GGUF

qwen35 · 4.7B · Q4_K_M

Not Rec.

Mar 5, 2026 · AMD RYZEN AI MAX+ 395

mistralai/devstral-small-2-2512

LM-STUDIO MLX

mistral3 · 24B · 4bit

Good

Mar 4, 2026 · Apple M4

Global Score
65 vs 69
Hardware Fit
73 vs 50
Quality Score
62 vs 77

Hardware

qwen3.5:4b mistralai/devstral…
MachineAZW GTR ProMacBook Air
CPUAMD RYZEN AI MAX+ 395Apple M4
Cores3210
RAM125 GB32 GB
GPUAMD Radeon 8060SApple M4
OSUbuntu 24.04.4 LTSmacOS 26.3
Archx64arm64
Power Modelow-powerbalanced

Performance

qwen3.5:4b mistralai/devstral…
Tokens/sec45.75.7
First chunk180 ms5007 ms
TTFT30.0 s5.0 s
Load time5.3 sN/A
Memory usage16.2 GB13.2 GB
Memory %13%41%

HW Fit Score Breakdown

qwen3.5:4b

Speed
43/50
TTFT
0/20
Memory
30/30

mistralai/devstral-smal…

Speed
11/50
TTFT
9/20
Memory
30/30

Quality

qwen3.5:4b

Reasoning
17/20
Coding
1/20
Instruction
13/20
Structured
8/15
Math
13/15
Multilingual
10/10
Reasoning: Strong Coding: Poor Instruction Following: Adequate Structured Output: Adequate Math: Strong Multilingual: Strong

mistralai/devstral-smal…

Reasoning
16/20
Coding
18/20
Instruction
13/20
Structured
14/15
Math
6/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Weak Multilingual: Strong

Run yours and compare

$ npm install -g metrillm@latest
$ metrillm

Requires Node 20+ and Ollama or LM Studio running

Or run without installing: npx metrillm@latest