Altman Says Codex 5.3 Outperforms Opus 4.6, as AI Coding Tools Enter a New Arms Race

Sam Altman publicly claimed OpenAI's Codex 5.3 is "much more effective" than Anthropic's Opus 4.6 for coding — a direct shot across the bow in the increasingly competitive AI developer tools market.

Sam Altman made a pointed comparison Thursday, stating that "Codex 5.3 is a much more effective model than Opus 4.6," as @sama posted to X. It's unusually direct competitive positioning from Altman, who typically frames OpenAI's advances in terms of capability rather than head-to-head comparisons with named competitors.

Unlock the full briefing

Get every story in today's briefing, the full archive, and the daily AI intelligence brief.

All stories today

Full archive

Daily brief

Cancel anytime. Payments powered by Stripe.