Back to leaderboard

codestral:22b

OLLAMA GGUF

llama · 22.2B · Q4_0

Good

Jun 26, 2026 · Cortex-X925

qwen3:14b

OLLAMA GGUF

qwen3 · 14.8B · Q4_K_M

Marginal

Mar 5, 2026 · Apple M1 Max

Global Score
67 vs 58
Hardware Fit
67 vs 82
Quality Score
67 vs 48

Hardware

codestral:22b qwen3:14b
MachineNVIDIA NVIDIA_DGX_SparkMacBook Pro
CPUCortex-X925Apple M1 Max
Cores2010
RAM122 GB32 GB
GPUDevice 2e12Apple M1 Max
OSUbuntu 24.04.4 LTSmacOS 26.3
Archarm64arm64
Power Modeperformancebalanced

Performance

codestral:22b qwen3:14b
Tokens/sec17.224.4
First chunk129 ms294 ms
TTFT129 ms7.9 s
Load time11.6 s3.7 s
Memory usage18.9 GB13.8 GB
Memory %16%43%

HW Fit Score Breakdown

codestral:22b

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

qwen3:14b

Speed
47/50
TTFT
5/20
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

qwen3:14b

Reasoning
13/20
Coding
6/20
Instruction
10/20
Structured
9/15
Math
2/15
Multilingual
8/10
Reasoning: Adequate Coding: Weak Instruction Following: Adequate Structured Output: Adequate Math: Poor 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