Kimi-48B:48B
OLLAMA GGUFkimi-linear · 49.1B · unknown
Aug 24, 2026 · Intel Core™ Ultra X7 358H
qwen3-vl-4b-instruct
LM-STUDIO GGUFqwen3vl · 4B · Q4_K_M
Apr 20, 2026 · Apple M2 Max
Global Score
86 vs 87
Hardware Fit
78 vs 100
Quality Score
89 vs 81
Hardware
Kimi-48B:48B qwen3-vl-4b-instruct
MachineGMKtec NucBox EVO-T2SMac Studio
CPUIntel Core™ Ultra X7 358HApple M2 Max
Cores1612
RAM64 GB96 GB
GPUIntel(R) Arc(TM) B390 GPUApple M2 Max
OSMicrosoft Windows 11 Pro 10.0.26200macOS 26.1
Archx64arm64
Power Modebalancedbalanced
Performance
Kimi-48B:48B qwen3-vl-4b-instruct
Tokens/sec23.171.2
First chunk752 ms4 ms
TTFT752 ms156 ms
Load time22.2 sN/A
Memory usage25.5 GB20.5 GB
Memory %40%21%
HW Fit Score Breakdown
Kimi-48B:48B
Speed
31/50
TTFT
20/20
Memory
27/30
qwen3-vl-4b-instruct
Speed
50/50
TTFT
20/20
Memory
30/30
Quality
Kimi-48B:48B
Reasoning
18/20
Coding
19/20
Instruction
15/20
Structured
15/15
Math
13/15
Multilingual
9/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Strong Multilingual: Strong
qwen3-vl-4b-instruct
Reasoning
15/20
Coding
17/20
Instruction
17/20
Structured
15/15
Math
8/15
Multilingual
9/10
Reasoning: Strong 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