baidu/ernie-4.5-21b-a3b
LM-STUDIO GGUFernie4_5-moe · 21B · Q4_K_M
Nov 1, 2026 · AMD RYZEN AI MAX+ 395
liquid/lfm2-24b-a2b
LM-STUDIO GGUFlfm2moe · 24B · Q4_K_M
Nov 1, 2026 · AMD RYZEN AI MAX+ 395
Global Score
86 vs 87
Hardware Fit
93 vs 100
Quality Score
83 vs 82
Hardware
baidu/ernie-4.5-21… liquid/lfm2-24b-a2b
MachineGMKtec EVO-X3GMKtec EVO-X3
CPUAMD RYZEN AI MAX+ 395AMD RYZEN AI MAX+ 395
Cores3232
RAM120 GB120 GB
GPUAMD Radeon(TM) 8060S GraphicsAMD Radeon(TM) 8060S Graphics
OSMicrosoft Windows 11 Pro 10.0.26300Microsoft Windows 11 Pro 10.0.26300
Archx64x64
Power Modebalancedbalanced
Performance
baidu/ernie-4.5-21… liquid/lfm2-24b-a2b
Tokens/sec46.380.9
First chunk33 ms7 ms
TTFT383 ms222 ms
Load time49.7 s26.6 s
Memory usage32.5 GB32.1 GB
Memory %27%27%
HW Fit Score Breakdown
baidu/ernie-4.5-21b-a3b
Speed
43/50
TTFT
20/20
Memory
30/30
liquid/lfm2-24b-a2b
Speed
50/50
TTFT
20/20
Memory
30/30
Quality
baidu/ernie-4.5-21b-a3b
Reasoning
16/20
Coding
17/20
Instruction
15/20
Structured
14/15
Math
12/15
Multilingual
9/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Strong Multilingual: Strong
liquid/lfm2-24b-a2b
Reasoning
15/20
Coding
17/20
Instruction
15/20
Structured
14/15
Math
11/15
Multilingual
10/10
Reasoning: Adequate Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Adequate 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