Back to leaderboard

google/gemma-3-4b

gemma3 · 4B · 4bit

Good

Mar 3, 2026 · Apple M4

qwen3.5:2b

qwen35 · 2.3B · Q8_0

Good

Mar 2, 2026 · Apple M4

Global Score
74 vs 73
Hardware Fit
93 vs 100
Quality Score
66 vs 62

Hardware

google/gemma-3-4b qwen3.5:2b
MachineMacBook AirMacBook Air
CPUApple M4Apple M4
Cores1010
RAM32 GB32 GB
GPUApple M4Apple M4
OSmacOS 26.3macOS 26.3
Archarm64arm64
Power Modebalancedbalanced

Performance

google/gemma-3-4b qwen3.5:2b
Tokens/sec22.230.4
TTFT1.9 s215 ms
Load time0.0 s1.8 s
Memory usage0.0 GB4.5 GB
Memory %0%14%

HW Fit Score Breakdown

google/gemma-3-4b

Speed
35/40
TTFT
28/30
Memory
30/30

qwen3.5:2b

Speed
40/40
TTFT
30/30
Memory
30/30

Quality

google/gemma-3-4b

Reasoning
10/20
Coding
17/20
Instruction
13/20
Structured
12/15
Math
4/15
Multilingual
10/10
Reasoning: Adequate Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Weak Multilingual: Strong

qwen3.5:2b

Reasoning
12/20
Coding
10/20
Instruction
13/20
Structured
14/15
Math
4/15
Multilingual
9/10
Reasoning: Adequate Coding: Weak Instruction Following: Adequate Structured Output: Strong Math: Weak Multilingual: Strong
View google/gemma-3-4b details | View qwen3.5:2b details

Run yours and compare

npx metrillm@latest bench

Requires Node 20+ and Ollama or LM Studio running