LLM Leaderboard
One page on 800+ language models: the best model for intelligence, coding, agents, and grunt work at a glance, then a sortable table and charts of Arena Elo, Artificial Analysis indexes, price, speed, and cost to run. Refreshed nightly.
Pulls current benchmark, pricing, and release data for every model on OpenRouter, LMArena, and Artificial Analysis, and renders it as one static page with no server: highlights above the fold, then a sortable table and charts in sections you can expand and collapse. Python stdlib only on the fetch side, vanilla JS on the page: no build step, no framework, no dependency to update. The sort, the filters, and the columns you turn on all live in the URL hash, so a link carries the exact view. A cron job on my VPS refetches every source nightly and pushes the pages here. It exists to answer one question I kept asking myself: which model is actually worth routing to right now.