GPT-6.2 scores worse than predecessors across the board on Company Agent Bench

philhchen · x · 2026-11-19

Josh Qin's team published evals showing GPT-6.2's Astra and Sol models scored lower than their predecessors on Company Agent Bench at every reasoning effort tested.

The takeaway: the upgrade is a regression for agentic workloads; full results are published.

Original post →

More from Models

Models channel →