gemma3n:latest
OLLAMA GGUFgemma3n · 6.9B · Q4_K_M
Mar 6, 2026 · AMD EPYC 9654P 96-Core Processor
lfm2.5:latest
OLLAMA GGUFlfm2moe · 8.5B · Q4_K_M
Jun 20, 2026 · AMD RYZEN AI MAX+ 395
Global Score
81 vs 79
Hardware Fit
100 vs 100
Quality Score
73 vs 70
Hardware
gemma3n:latest lfm2.5:latest
MachineSupermicro AS -2015HS-TNRAZW GTR Pro
CPUAMD EPYC 9654P 96-Core ProcessorAMD RYZEN AI MAX+ 395
Cores6432
RAM128 GB125 GB
GPUGH100 [H100L 94GB], ASPEED Graphics Family, GH100 [H100L 94GB]Radeon 8060S
OSUbuntu 24.04.3 LTSUbuntu 24.04.4 LTS
Archx64x64
Power Modeperformancelow-power
Performance
gemma3n:latest lfm2.5:latest
Tokens/sec108.7134.4
First chunk333 ms226 ms
TTFT333 ms226 ms
Load time4.5 s1.8 s
Memory usage8.6 GB6.9 GB
Memory %7%6%
HW Fit Score Breakdown
gemma3n:latest
Speed
50/50
TTFT
20/20
Memory
30/30
lfm2.5:latest
Speed
50/50
TTFT
20/20
Memory
30/30
Quality
gemma3n:latest
Reasoning
14/20
Coding
14/20
Instruction
14/20
Structured
14/15
Math
7/15
Multilingual
10/10
Reasoning: Adequate Coding: Adequate Instruction Following: Adequate Structured Output: Strong Math: Weak Multilingual: Strong
lfm2.5:latest
Reasoning
2/20
Coding
11/20
Instruction
17/20
Structured
15/15
Math
15/15
Multilingual
10/10
Reasoning: Poor Coding: Adequate 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