Lightweight open-source 20B model with reasoning support for efficient inference.
gpt-oss-20b
LLM Gateway routes requests to the best providers that are able to handle your prompt size and parameters.