Cloudflare serves Clef with SGLang: sub-40ms latency on a single H200
ying11231 · x · 2026-10-13
The SGLang team announced Cloudflare is now serving Clef with SGLang via its existing /v1/systemone API. The team upstreamed Clef support and performance optimizations: validated against the official reference implementation, sub-40ms decision latency for Clef-Flash on a single H200, and 64 questions answered in one prefill with text and image support. Clef-Omni support is next. Quoted context: new clef-omni takes audio, video, image, and text; clef-flash is now cheaper than jev; clef is 2x faster.
More from Infra
- Foxconn slides reveal Omniverse digital twins and GR00T humanoids building AI servers — firstadopter · 2026-10-13
- LiquidAI ships MLX versions of d1 models optimized for Apple Silicon local inference — JosephJacks_ · 2026-10-13
- Apple's new Private Cloud Compute release ships vserver images, hinting at virtualized macOS servers — jedisct1 · 2026-10-13
- HF's Zach Mueller: don't do this — most users can just rent or run sandboxes locally on CPUs — TheZachMueller · 2026-10-13
- Celesto CI undercuts GitHub Actions 9x: 1000 jobs for $6.67 vs $60 — aniketmaurya · 2026-10-13
- Keurig coffee maker flooded Wi-Fi with 1TB in 10 days — owner says unit bug — gnukeith · 2026-10-13