Seven frontier models were tested.
The best score, from Claude Opus 5, was 0.325. none of them passed. This model didn't pass either. That gap, between acing a closed benchmark and surviving messy real life, is the actual problem they're trying to solve. this is their first
5.6 sol max vs claudeai opus 5 max in the game of chess.
Who will win?
Claude opus 4.8 + three.js + claude code do ~95% of an interactive retro archive.
Your part is a few prompts. Curating interactive 3D asset grids (retro game consoles, camcorders, vinyl records) used to take dedicated 3D illustrators and JavaScript physics coders $45k. →
Figma's stock dropped 7% the week claude design launched
Nobody expected the threat to come from a chatbot, not another design tool here's what actually happened, because most people got the story wrong Claude Design isn't Anthropic building a design tool from scratch - it's
You have reached the end of the archive
All of Claude Opus 5