Someone Built a Fun Website That Simulates Your Machine's Tokens-Per-Second
RobertTAS · reddit · 2026-10-06
Reddit user RobertTAS shared a small toy website built out of boredom that simulates the tokens-per-second speed you'd get running LLMs locally. A lightweight novelty tool for roughly feeling out local inference performance differences.
More from Fun
- AI-generated video is so funny netizens say the compute was worth it — hexiang · 2026-10-06
- Bounty out for a working slopcannon Skyrim or Warcraft mod built with Claude — max_paperclips · 2026-10-06
- Rage at token waste: asking an LLM to decompile Linux, which is already open source — cyrus_zei · 2026-10-06
- Kolibri-1 plays Breakout with zero fine-tuning, 25ms per move — kmodi · 2026-10-06
- OpenAI's Dot agent sold a generator on Craigslist with just one prompt — andreisavu · 2026-10-06
- Suno recognized an electron wavefunction and wrote fitting lyrics for it — anderssandberg · 2026-10-06