OpenAI
gpt-oss-120b
Launched 2025-08-05 · Knowledge up to 2024-06 · Listed on OpenRouter 2025-08-05
Prices updated . Benchmarks updated .
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...
You send / 1M
$0.150
tokens in
It replies / 1M
$0.600
tokens out
Context
40K
tokens it can read at once
Typical request
$0.0005
1,000 words in, 400 out
Benchmarks
| Benchmark | Score | Kind | Date | Source |
|---|---|---|---|---|
| Epoch Capabilities Index | 139.8 | Independent | 2025-08-05 | Epoch AI |
| GPQA DiamondHigh effort | 75.8% | Independent | 2025-12-11 | Epoch AI |
| OTIS Mock AIMEHigh effort | 88.9% | Independent | 2025-12-11 | Epoch AI |
| Terminal-Benchagent Terminus 2 | 18.7% | Independent | 2025-11-01 | Terminal-Bench, republished by Epoch AI |
| CritPtHigh effort | 1.1% | Independent | 2025-08-05 | CritPt, republished by Epoch AI |
| Arena Elo (text) | 1,365 | Independent | 2026-10-28 | LMArena |
Rankings on this site use the Epoch Capabilities Index, an independent composite. Lab-reported rows are the lab's own figures and are kept separate.
Data privacy
No training or retention: Yes. At least one provider serves it with zero data retention.
| Provider | You send / 1M | It replies / 1M | Uptime | Privacy policy |
|---|---|---|---|---|
| PrimeIntellect · fp4 | $0.044 | $0.094 | 100% | Read |
| Near AI · fp4 | $0.030 | $0.150 | 100% | Read |
| Venice · int4 | $0.030 | $0.150 | 100% | Read |
| CoreWeave · fp4 | $0.030 | $0.170 | 99% | Read |
| DeepInfra · bf16 | $0.037 | $0.170 | 84% | Read |
| DekaLLM · bf16 | $0.030 | $0.180 | 99% | Read |
| AkashML · fp4 | $0.030 | $0.187 | 100% | Read |
| Crusoe · bf16 | $0.050 | $0.250 | 100% | Read |
| $0.090 | $0.360 | 100% | Read | |
| Mancer 2 · fp8 | $0.080 | $0.375 | 100% | Read |
| Sail Research · fp4 | $0.060 | $0.400 | 100% | Read |
| DigitalOcean | $0.060 | $0.420 | 100% | Read |
| BaseTen · fp4 | $0.100 | $0.500 | 99% | Read |
| BaseTen · fp4 | $0.100 | $0.500 | 100% | Read |
| Amazon Bedrock | $0.150 | $0.600 | 97% | Read |
| Nebius · fp4 | $0.150 | $0.600 | 100% | Read |
| Amazon Bedrock | $0.150 | $0.600 | 94% | Read |
| SiliconFlow · fp8 | $0.150 | $0.600 | 100% | Read |
| Together | $0.150 | $0.600 | 99% | Read |
| Groq | $0.150 | $0.600 | 100% | Read |
| Parasail · fp4 | $0.150 | $0.750 | 100% | Read |
| Cerebras · fp16 | $0.350 | $0.750 | 100% | Read |
| Parasail · fp4 | $0.100 | $0.750 | 100% | Read |
| SambaNova | $0.140 | $0.950 | 100% | Read |
Based on OpenRouter and Vercel AI Gateway data. Confirm terms with the provider before sending sensitive data.
What it can do
- Reads: text
- Writes: text
- Weights: Open weights
- Reasoning: Yes
- Tool calling: Yes
- Longest reply: 33K tokens
- Repeated text is billed at $0.075 per 1M (prompt caching)
Compare with
Appears in
Estimate your monthly cost with gpt-oss-120b
Sources and licenses
- OpenRouter: Prices, context windows, modalities and the model list come from the OpenRouter public API. Data: OpenRouter.
- Vercel AI Gateway: The no-training flag comes from the Vercel AI Gateway model list.
- models.dev: Lab launch dates, open-weights flags and knowledge cutoffs come from models.dev. Copyright (c) 2025 models.dev, MIT License.
- Epoch AI: Independent benchmark scores and the Epoch Capabilities Index (ECI) come from Epoch AI's Benchmarking Hub, used under CC BY 4.0. Epoch AI, Capabilities & benchmarking, epoch.ai/benchmarks. Scores Epoch marks as external keep their original license, and the source is named on each score.
- LMArena: Arena Elo (text, webdev and vision) comes from the LMArena leaderboard dataset (lmarena-ai/leaderboard-dataset), CC BY 4.0, leaderboard date 2026-10-28.
- Lab-reported scores, when an editor has added them, are the lab's own numbers. Each one links to the source and is labeled lab-reported. They are not used to rank the lists.
- What a score measures: the benchmark guide.
