We use cookies to measure how the site is used and how our ads perform (Google Analytics and Ads, Microsoft Clarity, and the Meta, OpenAI and X ad pixels). “Essentials only” turns off advertising cookies everywhere, and analytics too in the EEA, UK and Switzerland. Privacy Policy
KAIST AI's WikiTrace links every stored agent-memory lesson to the trajectory it came from, so agents can check lessons before acting. Memory-induced regression falls from 6.5% with unchecked lessons to 4.5%, while scores rise from 87.0 to 90.2.
Claude Fable 5.2 completes 41.3% of real paid Upwork projects on Scale Labs' Remote Labor Index — roughly twice Opus 5.5 and GPT-6 Astra, and far above the 2.5% best score at the benchmark's launch.
Qwen Code 1.0 Preview ships as a merged Qoder-plus-Qwen Code agent harness with native macOS, Linux and Windows packages that need no Node.js, plus session resumption and non-interactive runs for CI.
Cursor rebuilt Projects so teammates' coordinator agents share context and talk directly, flagging overlapping work before code reaches a pull request.