WHAT NEW IN X10.28 WED

AI collections from X · 22 collections · 58,925 posts

gpt-6IMAGE

Astra tops three computer-workflow benchmarks

Astra achieves state-of-the-art results on Agents' Last Exam, AutomationBench, and ScreenSpot Pro, benchmarks for computer workflow tasks across professions.

OpenAI
@OpenAIArchived 2026.09.13Originally posted 2026.09.04
Astra tops three computer-workflow benchmarks
OpenAI @OpenAIOriginally posted 2026.09.04
Open original on X

Astra achieves state-of-the-art results on Agents’ Last Exam, AutomationBench, and ScreenSpot Pro, benchmarks for computer workflow tasks across professions.

x.com/openai/status/2095595744300503356Archived 2026.09.13 13:28