ProductJun 11, 2026, 07:16 AM· United States
Surprise upset: GPT-5.5 beats Claude Fable 5 on brutal new Agents’ Last Exam benchmark
Summary
UC Berkeley's RDI and over 300 experts launched Agents’ Last Exam (ALE), a benchmark measuring AI's ability to execute economically valuable long-horizon workflows. OpenAI's GPT-5.5 topped the leaderboard with 24.0% pass rate, beating Anthropic's new Claude Fable 5 (22.0%). ALE aims to bridge academic benchmarks and real labor impact, showing top models still fundamentally fail.
Why it matters
This event shows the latest ranking shift in AI models on complex professional tasks, involving key companies OpenAI and Anthropic.
Source links
⚠ Content is from official & reputable-media public sources, AI-assisted and auto-published, for information only — not investment advice.
Market reaction
The following is market reference information for related companies, and does not constitute investment advice.
OpenAI
Private / not listed
Anthropic
Private / not listed