qwen3-30b-a3b-thinking-2507-claude-4.5-sonnet-high-reasoning-distill-mlx
LM-STUDIO MLXqwen3_moe · 30B · MXFP4
Apr 22, 2026 · Apple M2 Max
gemma4:31b
OLLAMA GGUFgemma4 · 31.3B · Q4_K_M
Jul 7, 2026 · Cortex-X925
Global Score
95 vs 83
Hardware Fit
94 vs 60
Quality Score
96 vs 93
Hardware
qwen3-30b-a3b-thin… gemma4:31b
MachineMac StudioASUSTeK COMPUTER INC. GX10
CPUApple M2 MaxCortex-X925
Cores1220
RAM96 GB122 GB
GPUApple M2 MaxDevice 2e12
OSmacOS 26.1Ubuntu 24.04.4 LTS
Archarm64arm64
Power Modebalancedperformance
Performance
qwen3-30b-a3b-thin… gemma4:31b
Tokens/sec80.410.9
First chunk46 ms681 ms
TTFT1.8 s681 ms
Load timeN/A16.6 s
Memory usage21.2 GB21.4 GB
Memory %22%18%
HW Fit Score Breakdown
qwen3-30b-a3b-thinking-…
Speed
50/50
TTFT
14/20
Memory
30/30
gemma4:31b
Speed
10/50
TTFT
20/20
Memory
30/30
Quality
qwen3-30b-a3b-thinking-…
Reasoning
19/20
Coding
18/20
Instruction
19/20
Structured
15/15
Math
15/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Strong Multilingual: Strong
gemma4:31b
Reasoning
19/20
Coding
19/20
Instruction
17/20
Structured
15/15
Math
14/15
Multilingual
9/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Strong Multilingual: Strong
Run yours and compare
$
npm install -g metrillm@latest$
metrillmRequires Node 20+ and Ollama or LM Studio running
Or run without installing: npx metrillm@latest