Show HN: I built a live race between LLMs to see which generates tokens fastestllm-racing.vercel.app 1 pointbaristaGeek6 months ago1 commentSaveHideCopy link On HNComments−baristaGeekOP6moI built this because I just thought it would be cool to show which LLMs respond faster (between GTP-4o mini, Claude 3 Haiku, and Gemini 2.5 Flash) and show some metrics (TTFT, avg tok/s, total time, and nTokens)
Comments
I built this because I just thought it would be cool to show which LLMs respond faster (between GTP-4o mini, Claude 3 Haiku, and Gemini 2.5 Flash) and show some metrics (TTFT, avg tok/s, total time, and nTokens)