OpenAI
OpenAI: gpt-oss-20b
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...
| Specifications | |
|---|---|
| Provider | OpenAI |
| Input pricing | $0.03/M tokens |
| Output pricing | $0.13/M tokens |
| Context length | 131K tokens |
| Max output | 131K tokens |
| Input modalities | text |
| Output modalities | text |
| Tokenizer | GPT |
| Knowledge cutoff | 2024-06-30 |
| Content moderated | No |
| HuggingFace | openai/gpt-oss-20b |
Supported Parameters
frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningreasoning_effortrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p