I think until Fable, Claude's vision was significantly worse than GPT and Gemini in my personal experience. I eval almost every vision model since I work on scr

I think until Fable, Claude's vision was significantly worse than GPT and Gemini in my personal experience. I eval almost every vision model since I work on screenshot to code conversion project: https://github.com/abi/screenshot-to-code https://github.com/abi/screenshot-to-code.