Back to leaderboard

qwen3:4b

OLLAMA GGUF

qwen3 · 4.0B · Q4_K_M

Not Rec.

Jun 26, 2026 · Cortex-A78AE

unsloth-phi-4

LM-STUDIO MLX

llama · 4bit

Excellent

Mar 6, 2026 · Apple M4 Pro

Global Score
32 vs 82
Hardware Fit
98 vs 88
Quality Score
4 vs 80

Hardware

qwen3:4b unsloth-phi-4
MachineNVIDIA NVIDIA Jetson Orin Nano Engineering Reference Developer Kit SuperMac mini
CPUCortex-A78AEApple M4 Pro
Cores614
RAM7 GB64 GB
GPUIntegrated / UnknownApple M4 Pro
OSUbuntu 22.04.5 LTSmacOS 15.7.4
Archarm64arm64
Power Modebalancedbalanced

Performance

qwen3:4b unsloth-phi-4
Tokens/sec17.927.4
First chunk950 ms457 ms
TTFT950 ms457 ms
Load time9.7 sN/A
Memory usage3.0 GB7.7 GB
Memory %40%12%

HW Fit Score Breakdown

qwen3:4b

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

unsloth-phi-4

Speed
38/50
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

unsloth-phi-4

Reasoning
15/20
Coding
19/20
Instruction
14/20
Structured
13/15
Math
9/15
Multilingual
10/10
Reasoning: Adequate Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Adequate 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