Agents' Last Exam
Tests agents on complex professional tasks in real software, from financial modelling to engineering and media production.
- Domain
- knowledge-work
- Published
- Sep 2026
- Notable for
- Headline professional-work benchmark at the GPT-6 Astra launch.
Cite
Notes
Only stored in your browser.
Top score 59.3% by GPT-6 Astra - 1 model reporting (1 frontier)
Top models
1
FAQ
- What is Agents' Last Exam?
- Tests agents on complex professional tasks in real software, from financial modelling to engineering and media production.
- What is the current top score on Agents' Last Exam?
- The top reported score is 59.3% by GPT-6 Astra, across 1 model reporting (1 from frontier labs).