gpt-oss-120b
OpenAI · Released Aug 2025
An open-weight 117-billion-parameter Mixture-of-Experts model from OpenAI with a 131k token context window.
- Strengths
- Handles complex reasoning and multi-step agentic tasks with a large effective parameter count through its sparse architecture.
- Best for
- Reasoning-heavy workloads and agentic systems where you need strong inference capabilities without the latency or cost of the full GPT-5 family.
- Limitations
- A larger open-weight model that requires more compute to run than gpt-oss-20b, and has a smaller context window than the GPT-5 family's 400k tokens.
Input / 1M
$0.037
Output / 1M
$0.17
Cached input / 1M
—
Context window
131K
Price history
Input (solid)Output (dashed)
Price change
- 30d
- in increased 23.3% out decreased 0.0%
- 90d
- in decreased 5.1% out decreased 5.6%
- 1y
- in decreased 5.1% out decreased 5.6%
- Since launch
- in decreased 5.1% out decreased 5.6%
Snapshots
| Effective | Input | Output | Cached in | Note | Source |
|---|---|---|---|---|---|
| 22 Aug 2026 | $0.037 | $0.17 | — | Imported from OpenRouter | openrouter.ai |
| 12 Aug 2026 | $0.03 | $0.17 | $0.03 | Imported from OpenRouter | openrouter.ai |
| 15 Jul 2026 | $0.037 | $0.17 | — | Imported from OpenRouter | openrouter.ai |
| 14 Jul 2026 | $0.03 | $0.15 | — | Imported from OpenRouter | openrouter.ai |
| 10 Jul 2026 | $0.036 | $0.18 | — | Imported from OpenRouter | openrouter.ai |
| 26 Jun 2026 | $0.03 | $0.15 | — | Imported from OpenRouter | openrouter.ai |
| 11 Jun 2026 | $0.039 | $0.18 | — | Imported from OpenRouter | openrouter.ai |