Benchmarking GPT‑OSS 120B: How OpenAI’s latest open‑weight model compares to GLM‑4.5, Qwen3, DeepSeek R1, and Kimi K2. Benchmarking GPT‑OSS 120B: How OpenAI’s latest open‑weight model compares to GLM‑4.5, Qwen3, DeepSeek R1, and Kimi K2.