Back to leaderboard

qwen3-4b-qwen3.6-plus-reasoning-distilled

LM-STUDIO GGUF

qwen3 · 4B · Q4_1

Good

Apr 21, 2026 · Apple M2 Max

codestral:22b

OLLAMA GGUF

llama · 22.2B · Q4_0

Good

Jun 26, 2026 · Cortex-X925

Global Score
65 vs 67
Hardware Fit
96 vs 67
Quality Score
51 vs 67

Hardware

qwen3-4b-qwen3.6-p… codestral:22b
MachineMac StudioNVIDIA NVIDIA_DGX_Spark
CPUApple M2 MaxCortex-X925
Cores1220
RAM96 GB122 GB
GPUApple M2 MaxDevice 2e12
OSmacOS 26.1Ubuntu 24.04.4 LTS
Archarm64arm64
Power Modebalancedperformance

Performance

qwen3-4b-qwen3.6-p… codestral:22b
Tokens/sec99.617.2
First chunk18 ms129 ms
TTFT1.6 s129 ms
Load time4.8 s11.6 s
Memory usage0.3 GB18.9 GB
Memory %0%16%

HW Fit Score Breakdown

qwen3-4b-qwen3.6-plus-r…

Speed
50/50
TTFT
16/20
Memory
30/30

codestral:22b

Speed
17/50
TTFT
20/20
Memory
30/30

Quality

qwen3-4b-qwen3.6-plus-r…

Reasoning
8/20
Coding
4/20
Instruction
14/20
Structured
9/15
Math
8/15
Multilingual
8/10
Reasoning: Weak Coding: Poor Instruction Following: Adequate Structured Output: Adequate Math: Adequate Multilingual: Strong

codestral:22b

Reasoning
11/20
Coding
17/20
Instruction
9/20
Structured
14/15
Math
6/15
Multilingual
10/10
Reasoning: Adequate Coding: Strong Instruction Following: Weak Structured Output: Strong Math: Weak Multilingual: Strong
View qwen3-4b-qwen3.6-p… details | View codestral:22b details

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