GPT-5.6 Pro beats Codex on critique and repair, says Will Depue
willdepue · x · 2026-07-24
Will Depue says GPT-5.6 Pro is substantially smarter than GPT-5.6 in Codex.
- He says Pro is especially good at critiquing and fixing the dumb ideas and implementations that Codex produces.
- The current workflow feels awkward: ask Codex to write a research report, paste it into Pro, then copy improved instructions back.
- His point is that there should be a better integrated way to move from agent-generated drafts to higher-quality critique and repair.
More from coding & agent
- OpenAI rolls out ChatGPT Voice on desktop with voice control for multiple agents — threepointone · 2026-07-24
- Y Combinator backs Daqstra, an AI-native layer for test infrastructure — ycombinator · 2026-07-24
- AI agents keep treating their own bad code as legacy and try to preserve it — cto_junior · 2026-07-24
- OpenAI Codex now supports one project across multiple folders — thesaraharminta · 2026-07-24
- The Ultimate Game for AI is the Articulation of Ideal State — JensHonack · 2026-07-24
- A new `/ce-handoff` skill moves agent context into a fresh session — kieranklaassen · 2026-07-24