How much does gpt-oss-120b (free) cost?

gpt-oss-120b (free) costs $0.00 per million input tokens and $0.00 per million output tokens. For a typical conversation (~2,000 tokens), that's approximately $0.000 per message.

What benchmarks has gpt-oss-120b (free) been tested on?

gpt-oss-120b (free) has been evaluated on 10 benchmarks. Top scores: OpenCompass — AIME2025: 93.4, OpenCompass — IFEval: 90.2, OpenCompass — MMLU-Pro: 79.7.

Is gpt-oss-120b (free) open source?

Yes, gpt-oss-120b (free) is open source.

How does gpt-oss-120b (free) compare to Gemini 2.5 Pro Preview 06-05?

gpt-oss-120b (free) has an average score of 74.2 while Gemini 2.5 Pro Preview 06-05 scores 73.5. gpt-oss-120b (free) outperforms Gemini 2.5 Pro Preview 06-05 overall. gpt-oss-120b (free) costs $0.00/1M input vs Gemini 2.5 Pro Preview 06-05 at $1.25/1M input. See full comparison →

Home/Models/gpt-oss-120b (free)

gpt-oss-120b (free)

Name: gpt-oss-120b (free)
Author: OpenAI

by OpenAI · Released Aug 2025

Open Source

74.2

avg score

Rank #31

Compare

Better than 87% of all models

Context

131K tokens (~66 books)

Input $/1M

Free

Output $/1M

Free

Type

text

License

Open Source

Benchmarks

10 tested

Data updated today

About

gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized...

Tested on 10 benchmarks with 68.7% average. Top scores: OpenCompass — AIME2025 (93.4%), OpenCompass — IFEval (90.2%), OpenCompass — MMLU-Pro (79.7%).

Capabilities

coding

60.1

#40 globally

math

93.4

#4 globally

knowledge

59.0

#54 globally

speed

55.4

#32 globally

language

90.2

#10 globally

Benchmark Scores

Compare All

Tested on 10 benchmarks · Ranked across 5 categories

Score Distribution (all 233 models)

0255075100

▲ You are here

codingCompare coding →

OpenCompass — LiveCodeBenchV6

OpenCompass Live Code Bench v6. Fresh competitive programming problems to evaluate code generation without memorization.

78.4—

Aider polyglot

Multi-language code editing from Aider. Tests editing ability across Python, JavaScript, TypeScript, Java, C++, Go, Rust, and more.

41.8—

mathCompare math →

OpenCompass — AIME2025

OpenCompass evaluation on AIME 2025 problems. Tests mathematical reasoning on fresh competition problems.

93.4—

knowledgeCompare knowledge →

OpenCompass — MMLU-Pro

OpenCompass MMLU-Pro evaluation. Harder knowledge test with more answer choices.

79.7—

OpenCompass — GPQA-Diamond

OpenCompass evaluation of GPQA Diamond. PhD-level science questions from the hardest subset.

78.9—

OpenCompass — HLE

OpenCompass evaluation of Humanitys Last Exam. Expert-level cross-discipline knowledge test.

18.3—

Quick compare:

vs Gemini 2.5 Pro Preview 06-05

vs Hermes 3 70B Instruct

vs Llama 3.3 70B Instruct

Excellent (85+) Good (70-85) Average (50-70) Below (<50)

Similar Models

Gemini 2.5 Pro Preview 06-05

Google DeepMind

73.5$1.25/1M

Hermes 3 70B Instruct

nousresearch

73.3$0.30/1M

Llama 3.3 70B Instruct

Frequently Asked Questions

gpt-oss-120b (free) is an open-source text AI model by OpenAI, released in August 2025. It has an average benchmark score of 74.2. Context window: 131K tokens.

Benchmarks

OpenCompass — AIME2025 OpenCompass — IFEval OpenCompass — MMLU-Pro OpenCompass — GPQA-Diamond OpenCompass — LiveCodeBenchV6

OpenAI · Provider OpenAI · Economy All Models Compare Models Pricing Developers · API Context Window · Glossary

gpt-oss-120b (free)

Frequently Asked Questions

Related Models

Benchmarks

Related Pages