We now have a new benchmark to compare LLMs. SVG Pokemons. It's obvious who's the best but which one do you think the worst? Prompt: Create an svg featuring 5 simplified pokemon characters arranged in a row: pikachu, bulbasaur, charmander, squirtle, eevee.

OMG Claude Sonnet 3.7 is insane. I got this stunning 3D city with one shot. Look at the shadows and the way the day transitions. This is just awesome.

I found a very simple way to pass terminal logs to Cursor using an MCP server which makes debugging easier. I'm writing all terminal output to the dev.log file. The MCP tool reads this file and passes its contents to Cursor so Agent can identify and fix the errors.

Grok 3 is out and I wanted to compare it with the other LLMs. I created a black hole simulation and although I'm not sure which version is the more realistic but I like Grok 3 better.

For a fun weekend project, I had Claude 3.5 Sonnet and GPT-4o play a game of Connect Four against each other. Guess who won every time?

Introducing AI Renamer for Desktop Last year I created an open source tool to rename files based on their content. But it was too complex for non-tech users. So I built a simple and easy-to-use desktop app to organize files with ease. Available for Mac and Windows. Link below.