SGLang v0.5.15 Boosts Production Inference Performance

SGLang released v0.5.15 with a focus on production inference optimizations, reporting significant serving speedups for GLM-5.2 NVFP4 on 8x B300 hardware.

2026-07-11 ~ 2026-07-12 · 2 related posts

1 near-duplicate retellings: BanghuaZ