New local AI benchmark site ranks models by hardware, counting prefill not just decode speed
draginol · x · 2026-11-06
A new Local AI benchmark website has launched, letting users compare LLM performance by both model and hardware. The author notes the speed score accounts for prefill as well as decode speed, making it more reflective of real-world local inference experience like time-to-first-token.
Related event: New Local AI Benchmark Site Compares 31 Models Across 15 Hardware Setups(2 posts)→
More from Infra
- SpaceX data center pledges 40% power discounts and billions in community benefits — Scobleizer · 2026-11-06
- Mystery seller seeks to dump $500 million of data center loans — dinabass · 2026-11-06
- Collinear AI runs full-scale agentic RL on Together Post-Training with self-onboarding — togethercompute · 2026-11-06
- SemiAnalysis: Google TPUv7 Delivers Up to 50% Better Performance-per-Dollar Than B200/B300 for Inference — vllm_project · 2026-11-06
- Meta President: new US data center build-outs will be 'very challenging' — SumitGup · 2026-11-06
- Veteran investor: memory stocks aren't cheap even at P/E of 3 as LTAs are meaningless — bookwormengr · 2026-11-06