baidu/ernie-4.5-21b-a3b
LM-STUDIO GGUFernie4_5-moe · 21B · Q4_K_M
Nov 1, 2026 · AMD RYZEN AI MAX+ 395
lmstudio-community/gemma-4-26b-a4b-it
LM-STUDIO GGUFgemma4 · 26B · Q4_K_M
Jun 29, 2026 · Intel Gen Intel® Core™ i9-11900H
Global Score
86 vs 81
Hardware Fit
93 vs 56
Quality Score
83 vs 91
Hardware
baidu/ernie-4.5-21… lmstudio-community…
MachineGMKtec EVO-X3Docker Container
CPUAMD RYZEN AI MAX+ 395Intel Gen Intel® Core™ i9-11900H
Cores3216
RAM120 GB15 GB
GPUAMD Radeon(TM) 8060S GraphicsTigerLake-H GT1 [UHD Graphics]
OSMicrosoft Windows 11 Pro 10.0.26300Ubuntu 26.04 LTS
Archx64x64
Power Modebalancedperformance
Performance
baidu/ernie-4.5-21… lmstudio-community…
Tokens/sec46.39.9
First chunk33 ms145 ms
TTFT383 ms4.4 s
Load time49.7 s48.3 s
Memory usage32.5 GB5.6 GB
Memory %27%37%
HW Fit Score Breakdown
baidu/ernie-4.5-21b-a3b
Speed
43/50
TTFT
20/20
Memory
30/30
lmstudio-community/gemm…
Speed
18/50
TTFT
10/20
Memory
28/30
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
lmstudio-community/gemm…
Reasoning
17/20
Coding
19/20
Instruction
18/20
Structured
14/15
Math
13/15
Multilingual
10/10
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