WhatLLM.org

Hey everyone! 👋 I'm Dylan, and I built WhatLLM because I was tired of endlessly searching "GPT vs Claude vs Gemini" every time a new model dropped. **The problem**: There are now 200+ LLMs, and picking one is overwhelming. Benchmarks are scattered. Pricing varies 10x between providers for the *same* model. Speed matters for production but nobody talks about it. **My solution**: One dashboard with everything, quality scores, prices, speed, context windows, filterable by use case. **Some things I learned building this:** - DeepSeek R1 can be $0.55/M tokens or $7/M depending on provider - Gemini 3 Pro just tied GPT-5.2 on quality (at 1/3 the price) **What would make WhatLLM more useful for you?** I'm actively adding features and would love your feedback! P.S. The data mostly comes from Artificial Analysis, the gold standard for LLM benchmarks. Updated weekly.