OpenAI: gpt-oss-20bopenai
Bottom Line
OpenAI: gpt-oss-20b: $0.02/1M input, $0.09/1M output, 131K context, CPS grade A. gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for...
API Pricing (per 1M tokens)
| Input | $0.02 |
| Output | $0.09 |
| Cache Read | — |
| Cache Write | — |
| Blended (3:1) | $0.04 |
Specs & Performance
| Context Length | 131K |
| Max Output | 33K |
| Intelligence | 9 |
| Coding | 21 |
| Math | 89 |
| Speed | 160 tok/s |
| TTFT | 0.45s |
| CPS Grade | A |
| Input / Output Modalities | text → text |
More from openai
Looking for cheaper alternatives?
Find the cheapest model by task, or compare two models head-to-head.
FAQ
How much does the OpenAI: gpt-oss-20b API cost?
OpenAI: gpt-oss-20b costs $0.02/1M input tokens and $0.09/1M output tokens. Blended cost at a 3:1 ratio is about $0.04/1M tokens.
What is the context window of OpenAI: gpt-oss-20b?
OpenAI: gpt-oss-20b supports a 131K token context window with up to 33K output tokens.
Is OpenAI: gpt-oss-20b worth it?
OpenAI: gpt-oss-20b has a CPS grade of A (score 65), intelligence index 9, coding index 21. Use the compare page to benchmark against rivals.