## LLM Rankings ![[Screenshot 2025-05-12 at 08.08.16.png]] I love [OpenRouter rankings](https://openrouter.ai/rankings) - these are users voting with their dollars, working on real world use case, and is likely a good proxy for which model is leading. The chart clearly shows when the world had woken up to LLMs (late Jan 2025) - when DeepSeek, a lesser known Chinese AI trading shop turned AI player, launched their chatbot, it surpassed ChatGPT as the most downloaded freeware app on the iOS App Store in the United States. This triggered an 18% drop in Nvidia's share price. When I am 12 hours deep into programming, I stare at the screen as if staring into the abyss - code output quality and coherence drops dramatically. Using an LLM feels like going to the gym where you are on the final rep, and pushing until you can't push anymore, and at this point, it helps you finish the rep and more. (You still have to go to the gym and put in the work, though - no amount of automation is going to save you from that). [OpenTools](https://opentools.com/): Connect LLM to tools and data through a standard interface