GeekRadar / The comparison desk

COMPARE.

AI models

openai

GPT OSS 120B

AI model Released Aug 5, 2025

Context window: 131,072 tokens
Maximum output: 32,768 tokens
Open weights: Yes

Capability & preference

General capability · ECI

A cross-benchmark capability estimate. Index points, not a percentage; intervals show uncertainty.

GPT OSS 120B 139.8^* gpt-oss-120b · 95% interval 135.6–142.2

Human preference · Arena

Overall text preferences from head-to-head votes. Not factual accuracy or a task-success rate.

GPT OSS 120B 1365.5 gpt-oss-120b · 95% interval 1361.1–1369.8 · 30,200 votes · 2026-10-28

Model performance

Radar task score: four tests, equal weight. An experimental summary of these tasks, not a universal intelligence rating.

GPT OSS 120B

Tested configuration Radar task score

Unrated

1/4 core tests · not enough evidence

1 benchmark results for this configuration

Scientific reasoning

GPQA Diamond · 25%

Expert-written science questions; strict multiple-choice scoring. Chance accuracy is 25%.

GPT OSS 120B Not available

Novel problem solving

Mystery Game Puzzles · 25%

Best-next-move puzzles in an undisclosed game variant. A narrow reasoning test, not general intelligence.

GPT OSS 120B 0.0^*% Uncertainty not available · 2026-08-27

Advanced mathematics

FrontierMath Tiers 1-3 · 25%

Private v2 mathematics problems with Python tools. Earlier versions and Tier 4 are excluded.

GPT OSS 120B Not available

Factual answers

SimpleQA Verified · 25%

Correct factual answers under a guessing prompt, graded by a model. Not a hallucination or abstention score.

GPT OSS 120B Not available

Bars show reported task scores. Whiskers are approximate 95% intervals where standard errors are available. Different reasoning budgets are not equal-cost tests. Missing results are never zero.

Recorded results

Only the same benchmark, version, workload, scenario and unit share a row. Missing results are not zero. System results include the full submitted hardware and software configuration.

Arena text preference / chinesehuman preference · 2026-10-28 · rating

GPT OSS 120B

1,384.78 rating

Confidence interval: 1,372.21 – 1,397.36

2275 votes

Oct 28, 2026

Category: chinese
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / codinghuman preference · 2026-10-28 · rating

GPT OSS 120B

1,379.23 rating

Confidence interval: 1,371.51 – 1,386.94

6405 votes

Oct 28, 2026

Category: coding
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / creative_writinghuman preference · 2026-10-28 · rating

GPT OSS 120B

1,274.49 rating

Confidence interval: 1,264.78 – 1,284.21

3908 votes

Oct 28, 2026

Category: creative_writing
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / englishhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,374.63 rating

Confidence interval: 1,368.77 – 1,380.49

13026 votes

Oct 28, 2026

Category: english
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating

GPT OSS 120B

1,330.38 rating

Confidence interval: 1,324.38 – 1,336.39

21443 votes

Oct 28, 2026

Category: exclude_ties
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / experthuman preference · 2026-10-28 · rating

GPT OSS 120B

1,356.03 rating

Confidence interval: 1,340.29 – 1,371.77

1401 votes

Oct 28, 2026

Category: expert
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / frenchhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,369.82 rating

Confidence interval: 1,341.89 – 1,397.75

461 votes

Oct 28, 2026

Category: french
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / germanhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,353.15 rating

Confidence interval: 1,330.43 – 1,375.87

672 votes

Oct 28, 2026

Category: german
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / hard_promptshuman preference · 2026-10-28 · rating

GPT OSS 120B

1,363.33 rating

Confidence interval: 1,357.70 – 1,368.95

14681 votes

Oct 28, 2026

Category: hard_prompts
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,367.98 rating

Confidence interval: 1,360.30 – 1,375.66

6480 votes

Oct 28, 2026

Category: hard_prompts_english
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating

GPT OSS 120B

1,358.86 rating

Confidence interval: 1,350.50 – 1,367.23

5358 votes

Oct 28, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating

GPT OSS 120B

1,292.28 rating

Confidence interval: 1,284.17 – 1,300.39

5494 votes

Oct 28, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating

GPT OSS 120B

1,354.61 rating

Confidence interval: 1,340.90 – 1,368.33

1875 votes

Oct 28, 2026

Category: industry_legal_and_government
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating

GPT OSS 120B

1,381.51 rating

Confidence interval: 1,372.73 – 1,390.28

4826 votes

Oct 28, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,386.06 rating

Confidence interval: 1,371.37 – 1,400.75

1547 votes

Oct 28, 2026

Category: industry_mathematical
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating

GPT OSS 120B

1,387.59 rating

Confidence interval: 1,372.61 – 1,402.57

1627 votes

Oct 28, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating

GPT OSS 120B

1,387.34 rating

Confidence interval: 1,380.93 – 1,393.74

10214 votes

Oct 28, 2026

Category: industry_software_and_it_services
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating

GPT OSS 120B

1,305.47 rating

Confidence interval: 1,297.90 – 1,313.03

6646 votes

Oct 28, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating

GPT OSS 120B

1,317.84 rating

Confidence interval: 1,310.82 – 1,324.86

7863 votes

Oct 28, 2026

Category: instruction_following
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / japanesehuman preference · 2026-10-28 · rating

GPT OSS 120B

1,331.02 rating

Confidence interval: 1,305.70 – 1,356.35

578 votes

Oct 28, 2026

Category: japanese
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / koreanhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,282.15 rating

Confidence interval: 1,258.44 – 1,305.87

681 votes

Oct 28, 2026

Category: korean
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / longer_queryhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,318.98 rating

Confidence interval: 1,311.33 – 1,326.63

6559 votes

Oct 28, 2026

Category: longer_query
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / mathhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,388.34 rating

Confidence interval: 1,374.54 – 1,402.14

1778 votes

Oct 28, 2026

Category: math
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / multi_turnhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,339.63 rating

Confidence interval: 1,330.90 – 1,348.37

4880 votes

Oct 28, 2026

Category: multi_turn
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / non_englishhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,350.42 rating

Confidence interval: 1,345.07 – 1,355.76

17135 votes

Oct 28, 2026

Category: non_english
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / overallhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,365.46 rating

Confidence interval: 1,361.11 – 1,369.80

30200 votes

Oct 28, 2026

Category: overall
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / polishhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,353.08 rating

Confidence interval: 1,338.57 – 1,367.59

1732 votes

Oct 28, 2026

Category: polish
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / russianhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,342.55 rating

Confidence interval: 1,330.06 – 1,355.04

2138 votes

Oct 28, 2026

Category: russian
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Arena text preference / spanishhuman preference · 2026-10-28 · rating

GPT OSS 120B

1,388.84 rating

Confidence interval: 1,367.64 – 1,410.04

855 votes

Oct 28, 2026

Category: spanish
Evaluated configuration: gpt-oss-120b
Evaluation type: human preference

Run it locally

Inference · memory estimate

GPT OSS 120B

Not enough verified architecture information for a reliable VRAM estimate.

API cost

USD / month

Input tokens / monthOutput tokens / month

Standard text-token estimate. Excludes taxes, caching, batch discounts, reasoning and multimodal surcharges. Provider limits and regional availability may differ.

GPT OSS 120B

Provider offer

GPT OSS 120B / Pricing tiers 5

Provider / scope	Rate	Currency	Amount / unit
AWS Bedrockgpt-oss-120bStandard	Input	USD	0.15 / 1M tokens
AWS Bedrockgpt-oss-120bStandard Global cross-region	Input	USD	0.1545 / 1M tokens
AWS Bedrockgpt-oss-120bStandard Global cross-region	Output	USD	0.618 / 1M tokens
DeepInfraopenai/gpt-oss-120bStandard	Input	USD	0.037 / 1M tokens
DeepInfraopenai/gpt-oss-120bStandard	Output	USD	0.17 / 1M tokens
GroqGPT OSS 120BStandard	Input	USD	0.15 / 1M tokens
GroqGPT OSS 120BStandard	Output	USD	0.6 / 1M tokens
SambaNova Cloudgpt-oss-120bStandard	Input	USD	0.22 / 1M tokens
SambaNova Cloudgpt-oss-120bStandard	Output	USD	0.59 / 1M tokens

Capabilities

Context window: 131,072 tokens
Maximum output: 32,768 tokens
Reasoning: Yes
Tool calling: Yes
Open weights: Yes
Knowledge cutoff: 2024-06

Input modalities

text

Output modalities

text

Processing & platform

Architecture: gpt_oss

* Provisional data

Capability & preference

General capability · ECI

Human preference · Arena

Model performance

GPT OSS 120B

Scientific reasoning

Novel problem solving

Advanced mathematics

Factual answers

Competition mathematics

Recorded results

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

GPT OSS 120B

Run it locally

GPT OSS 120B

API cost

GPT OSS 120B

Capabilities

Input modalities

Output modalities

Processing & platform