Superpower Daily2 min read
GPT-6 Astra leads coding test but finishes only 26.4% of tasks, The Decoder reports
The November 8 coverage identifies the leading model in Co-Design Bench and details how changing its working environment can alter scores. The study deliberately favors difficult requests.
By Superpower DailySuperpower Daily

Read the full story
Originally published by Superpower Daily on Nov 8, 2026.
Comments
Sign in to join the discussion.