GPT-120B:120b

gptoss · 116.8B · MXFP4

THINKING MODEL

HP OMEN 35L VALORANT Limited Edition Gaming Desktop PC (Intel Core™ i7-14700F)

64 GB · Microsoft Windows 11 Enterprise 10.0.26200

Tested on August 18, 2026

Top 69% Compare

Global Score

66 /100

Not Rec.

Hardware Fit

14/100

Quality

88/100

Get this model

🦙

Ollama

ollama pull GPT-120B:120b

View on Ollama Library

ollama.com/library/GPT-120B

Get it in LM Studio

Search and download models directly from the app

🤗

Find on HuggingFace

GGUF versions & quantizations

Hardware

Machine: HP OMEN 35L VALORANT Limited Edition Gaming Desktop PC
CPU: Intel Core™ i7-14700F
Cores: 28 total (28 perf)
Frequency: 2.1 GHz
RAM: 64 GB DDR5
GPU: NVIDIA GeForce RTX 5080
OS: Microsoft Windows 11 Enterprise 10.0.26200
Arch: x64
Power Mode: balanced

Performance

Tokens/sec Tokens generated per second — higher is better: 10.1
Standard deviation Variation of speed across benchmark runs — lower means more consistent: ±1.0
First chunk latency Delay before the first streamed chunk arrives from runtime: 6.6 s
Time to first token How fast the model starts responding — lower is better: 10.3 s
Load time Time to load the model into memory before inference: 49.8 s
Memory usage RAM consumed during inference — percentage of total system memory: 61.5 GB (96%)
Total tokens Total number of tokens generated across all benchmark prompts: 1485
Thinking tokens (est.) Estimated internal reasoning tokens — models with "thinking" generate more tokens: ~271

Score breakdown

Speed

10/50

Time to first token

2/20

Memory

2/30

Quality

Reasoning

18/20

Coding

19/20

Instruction following

11/20

Structured output

15/15

Math

15/15

Multilingual

10/10

Category levels

Reasoning: Strong Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Strong Multilingual: Strong

Metadata

Spec version: 0.2.1
Runtime: Ollama 0.32.14
Model format: GGUF
Hardware profile: HIGH-END
Result hash: dd2578e0b1297c15d50463a307faeb8549d660d40bb6ee1f2ebbecc2b9dbe631

Interpretation

Hardware fit: 14/100. Overall suitability: NOT RECOMMENDED (Global 66/100). Category profile: Reasoning: Strong, Coding: Strong, Instruction Following: Adequate, Structured Output: Strong, Math: Strong, Multilingual: Strong.

Warnings

Significant swap activity during benchmark (+1.0 GB). Model may exceed available RAM — results are severely degraded.

Disqualifiers

Memory usage critical: model delta +96%

Bench Environment

Swap delta: +1.0 GB CPU load: avg 54% (peak 77%)

Run yours now

$ npm install -g metrillm@latest

$ metrillm

Requires Node 20+ and Ollama or LM Studio running

Or run without installing: npx metrillm@latest