Back to leaderboard

codestral:22b

OLLAMA GGUF

llama · 22.2B · Q4_0

Good

Jun 26, 2026 · Cortex-X925

qwen3.5-27b

LM-STUDIO MLX

qwen3_5 · 27B · 4bit

Good

Mar 3, 2026 · Apple M4

Global Score
67 vs 75
Hardware Fit
67 vs 54
Quality Score
67 vs 84

Hardware

codestral:22b qwen3.5-27b
MachineNVIDIA NVIDIA_DGX_SparkMacBook Air
CPUCortex-X925Apple M4
Cores2010
RAM122 GB32 GB
GPUDevice 2e12Apple M4
OSUbuntu 24.04.4 LTSmacOS 26.3
Archarm64arm64
Power Modeperformancebalanced

Performance

codestral:22b qwen3.5-27b
Tokens/sec17.26.6
First chunk129 ms4364 ms
TTFT129 ms4.4 s
Load time11.6 sN/A
Memory usage18.9 GB15.0 GB
Memory %16%47%

HW Fit Score Breakdown

codestral:22b

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

qwen3.5-27b

Speed
13/50
TTFT
11/20
Memory
30/30

Quality

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

qwen3.5-27b

Reasoning
17/20
Coding
18/20
Instruction
11/20
Structured
15/15
Math
13/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Strong 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