Back to leaderboard

qwen3:4b

OLLAMA GGUF

qwen3 · 4.0B · Q4_K_M

Not Rec.

Jun 26, 2026 · Cortex-A78AE

qwen/qwen3.8-27b

LM-STUDIO MLX

qwen3_5 · 27B · 8bit

Excellent

Aug 18, 2026 · Apple M4 Max

Global Score
32 vs 81
Hardware Fit
98 vs 68
Quality Score
4 vs 87

Hardware

qwen3:4b qwen/qwen3.8-27b
MachineNVIDIA NVIDIA Jetson Orin Nano Engineering Reference Developer Kit SuperMacBook Pro
CPUCortex-A78AEApple M4 Max
Cores616
RAM7 GB128 GB
GPUIntegrated / UnknownApple M4 Max
OSUbuntu 22.04.5 LTSmacOS 26.5.2
Archarm64arm64
Power Modebalancedbalanced

Performance

qwen3:4b qwen/qwen3.8-27b
Tokens/sec17.916.4
First chunk950 ms5 ms
TTFT950 ms742 ms
Load time9.7 sN/A
Memory usage3.0 GB38.5 GB
Memory %40%30%

HW Fit Score Breakdown

qwen3:4b

Speed
50/50
TTFT
20/20
Memory
28/30

qwen/qwen3.8-27b

Speed
18/50
TTFT
20/20
Memory
30/30

Quality

qwen3:4b

Reasoning
4/20
Coding
0/20
Instruction
0/20
Structured
0/15
Math
0/15
Multilingual
0/10
Reasoning: Poor Coding: Poor Instruction Following: Poor Structured Output: Poor Math: Poor Multilingual: Poor

qwen/qwen3.8-27b

Reasoning
17/20
Coding
18/20
Instruction
16/20
Structured
15/15
Math
11/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Adequate Multilingual: Strong
View qwen3:4b details | View qwen/qwen3.8-27b 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