Model by openai

gpt-oss-120b

Mixture of Experts (MoE) reasoning LLM (text-only) designed to fit within 80GB GPU.