GLM-5.2-NVFP4-REAP-469B serving on SM120 (4× RTX PRO 6000 Blackwell) — one-command vLLM launch recipe, 250K context, DeepSeek Sparse Attention + MTP speculative decode

15211Shellnew✓ Reviewed

Roles are derived from products, launches and repositories we track — not from self-description.