CMU and Browser Use release Odysseys2, where SOTA web agents pass only 47/100

CMU and Browser Use released Odysseys2, a harder long-horizon web agent benchmark; the best SOTA agent completed only 47 of 100 tasks, and five agents failed entirely.

2026-10-28 ~ 2026-10-29 · 3 related posts

• Odysseys2: CMU and Browser Use's brutal web-agent benchmark where SOTA completes just 47/100 — dan_fried · 2026-10-28
• CMU and Browser Use launch Odysseys2: best agent completes only 47/100 real web tasks — wellecks · 2026-10-28
• Odysseys2: new live-web benchmark where SOTA agent completes only 47/100 tasks — kohjingyu · 2026-10-29