Sublime

An inspiration engine for ideas

AllPeopleCollectionsArticlesAudioBooksFilesHighlightsImagesLinksNotesTextTweetsVideosSocial

We now have an open-source model that is beating Claude 3 Opus... being served at nearly **300 tokens per second** on @GroqInc. The applications built off of this tech will be nothing short of revolutionary.

Matt Shumer

x.com

We need to talk about this tool… 87 images in 1.82s. That’s 40x faster than @vercel's next/og. https://t.co/WmLtB4ymTs

Christopher Burns

x.com

Fastest inference engine for LLMs! LMCache is an LLM serving engine that reduce Time to First Token (TTFT) and increase throughput, especially under long-context scenarios. 100% Open Source https://t.co/xHAFz7d9v8

Sumanth

x.com

https://t.co/TwLXcXdAtE

9999

x.com

Llama 3.3 70B at 275 tps and this pricing is so good. I'm moving all of my structured output queries to it now. https://t.co/Q1xdSK8ecb

Shayan

x.com

I’m doing performance testing of my system at Bloomberg. A single request goes through 4 microservices, all written in OCaml. It takes 10ms total to process it.

Dmitrii Kovanikov x.com

@DylanoA4 https://t.co/gZjDTMkPgz

jack

x.com

bruh in just one day - over 1,400+ users - over 180 projects created - over 273 agents initialized - over 100 videos processed

Micky x.com

woooaw 😮zustand + mmkv + tanstack hitting <1.4ms cache restore on real device Every time I relaunch the app, my whole session & state are back — instantly. No loaders. No flickers. Just fast ⚡️ This @expo boilerplate is getting scary good 🧠🔥 https://t.co/yvNCEmh0QF

x.com