Back to leaderboard

qwen3:30b-a3b

OLLAMA GGUF

qwen3moe · 30.5B · Q4_K_M

Excellent

Jun 25, 2026 · Cortex-X925

qwen2.5:1.5b

OLLAMA GGUF

qwen2 · 1.5B · Q4_K_M

Good

Jun 17, 2026 · ZHAOXIN KaiXian KX-7000 KaiXian KX-7000

Global Score
84 vs 65
Hardware Fit
93 vs 87
Quality Score
80 vs 56

Hardware

qwen3:30b-a3b qwen2.5:1.5b
MachineNVIDIA NVIDIA_DGX_SparkShanghai Zhaoxin Semiconductor Co., Ltd. ZXE CRB
CPUCortex-X925ZHAOXIN KaiXian KX-7000 KaiXian KX-7000
Cores208
RAM122 GB30 GB
GPUDevice 2e12ZX C-1190
OSUbuntu 24.04.4 LTSMicrosoft Windows 11 Pro 10.0.26100
Archarm64x64
Power Modeperformancebalanced

Performance

qwen3:30b-a3b qwen2.5:1.5b
Tokens/sec92.716.8
First chunk286 ms941 ms
TTFT1.5 s941 ms
Load time15.3 s0.2 s
Memory usage20.3 GB1.0 GB
Memory %17%3%

HW Fit Score Breakdown

qwen3:30b-a3b

Speed
50/50
TTFT
13/20
Memory
30/30

qwen2.5:1.5b

Speed
37/50
TTFT
20/20
Memory
30/30

Quality

qwen3:30b-a3b

Reasoning
16/20
Coding
10/20
Instruction
18/20
Structured
13/15
Math
13/15
Multilingual
10/10
Reasoning: Strong Coding: Adequate Instruction Following: Strong Structured Output: Strong Math: Strong Multilingual: Strong

qwen2.5:1.5b

Reasoning
8/20
Coding
10/20
Instruction
12/20
Structured
13/15
Math
5/15
Multilingual
8/10
Reasoning: Weak Coding: Adequate Instruction Following: Adequate Structured Output: Strong Math: Weak Multilingual: Strong
View qwen3:30b-a3b details | View qwen2.5:1.5b 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