liquid/lfm2-24b-a2b
LM-STUDIO GGUFlfm2moe · 24B · Q4_K_M
Nov 1, 2026 · AMD RYZEN AI MAX+ 395
granite3.1-dense:2b
OLLAMA GGUFgranite · 2.5B · Q4_K_M
Mar 2, 2026 · Apple M4
Global Score
87 vs 68
Hardware Fit
100 vs 100
Quality Score
82 vs 54
Hardware
liquid/lfm2-24b-a2b granite3.1-dense:2b
MachineGMKtec EVO-X3MacBook Air
CPUAMD RYZEN AI MAX+ 395Apple M4
Cores3210
RAM120 GB32 GB
GPUAMD Radeon(TM) 8060S GraphicsApple M4
OSMicrosoft Windows 11 Pro 10.0.26300macOS 26.3
Archx64arm64
Power Modebalancedbalanced
Performance
liquid/lfm2-24b-a2b granite3.1-dense:2b
Tokens/sec80.954.6
First chunk7 msN/A
TTFT222 ms121 ms
Load time26.6 s0.4 s
Memory usage32.1 GB5.6 GB
Memory %27%18%
HW Fit Score Breakdown
liquid/lfm2-24b-a2b
Speed
50/50
TTFT
20/20
Memory
30/30
granite3.1-dense:2b
Speed
40/50
TTFT
30/20
Memory
30/30
Quality
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
granite3.1-dense:2b
Reasoning
8/20
Coding
9/20
Instruction
13/20
Structured
14/15
Math
3/15
Multilingual
7/10
Reasoning: Weak Coding: Weak Instruction Following: Adequate Structured Output: Strong Math: Poor Multilingual: Adequate
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