Back to leaderboard

hf.co/InternScience/Agents-A1-4B-Q8_0-GGUF:Q8_0

OLLAMA GGUF

qwen35 · 4.21B · unknown

Marginal

Aug 22, 2026 · Apple M1 Pro

GPT-20B:20b

OLLAMA GGUF

gptoss · 20.9B · MXFP4

Excellent

Aug 22, 2026 · NVIDIA RTX Spark N1X (6144-core GPU, 20-core ) RTX Spark N1X (6144-core GPU, 20-core )

Global Score
59 vs 93
Hardware Fit
97 vs 91
Quality Score
43 vs 94

Hardware

hf.co/InternScienc… GPT-20B:20b
MachineMacBook ProHP HyperX OMEN Transcend Gaming Laptop 14-ga0xxx5CD612DXML
CPUApple M1 ProNVIDIA RTX Spark N1X (6144-core GPU, 20-core ) RTX Spark N1X (6144-core GPU, 20-core )
Cores1020
RAM16 GB55 GB
GPUApple M1 ProNVIDIA RTX Spark N1X (6144-core Blackwell RTX GPU), NVIDIA NPU
OSmacOS 26.6.2Microsoft Windows 11 Pro 10.0.28000
Archarm64arm64
Power Modebalancedbalanced

Performance

hf.co/InternScienc… GPT-20B:20b
Tokens/sec30.957.8
First chunk891 ms203 ms
TTFT891 ms3.0 s
Load time3.6 s19.2 s
Memory usage6.6 GB12.0 GB
Memory %41%22%

HW Fit Score Breakdown

hf.co/InternScience/Age…

Speed
50/50
TTFT
20/20
Memory
27/30

GPT-20B:20b

Speed
50/50
TTFT
11/20
Memory
30/30

Quality

hf.co/InternScience/Age…

Reasoning
13/20
Coding
15/20
Instruction
2/20
Structured
0/15
Math
13/15
Multilingual
0/10
Reasoning: Adequate Coding: Adequate Instruction Following: Poor Structured Output: Poor Math: Strong Multilingual: Poor

GPT-20B:20b

Reasoning
19/20
Coding
17/20
Instruction
18/20
Structured
15/15
Math
15/15
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
$ metrillm

Requires Node 20+ and Ollama or LM Studio running

Or run without installing: npx metrillm@latest