LLMOpen weights
gpt-oss-120b
OpenAI
Updated May 21, 2026 · Benchmarks via Artificial Analysis, specs & pricing via OpenRouter · Methodology · Report an error
Specifications
- Type
- LLM
- Access
- Open weights
- Released
- August 5, 2025
- License
- Apache 2.0
- Parameters
- 117B (5.1B active)
- Context window
- 131,072 tokens
- Knowledge cutoff
- June 30, 2024
- Output speed
- 500 tok/s
- Latency (TTFT)
- 0.5s
- Input
- Text
- Output
- Text
- API pricing
- $0.04 in · $0.18 out / 1M tokens
Capabilities
Function callingStructured outputCode execution
Benchmarks
Reasoning
| GPQA Diamond | 80.9 |
Coding
| SWE-bench Verified | 62.4 |
| LiveCodeBench | 87.8 |
Math
| AIME 2025 | 93.4 |
Agents
| TAU-bench Retail | 67.8 |
General
| Humanity’s Last Exam | 19 |
| MMLU | 90 |
| MMLU-Pro | 80.8 |
Our take
OpenAI’s first open-weights release since GPT-2 — an MoE reasoning model that runs on a single 80GB GPU and approaches o4-mini on core reasoning.