GPT-6.2 Sol drops 3 points on Coding Agent Index despite release hype

ArtificialAnlys · x · 2026-11-18

Artificial Analysis notes that on the Coding Agent Index (Codex harness, max effort), GPT-6.2 Astra is flat with GPT-6 Astra while GPT-6.2 Sol loses 3 points vs GPT-6.1 Sol — driven by a 14-point drop on SWE-Atlas-QnA-AA, with the new model more likely to omit key facts. Several effort tiers still sit on the Cost-per-Task Pareto frontier; Claude Code variants still top the index.

Related event: GPT-6.2 Cuts Price 20% but Coding Stalls in Third-Party Benchmarks(4 posts)→

Original post →

More from coding & agent

coding & agent channel →