@gdb: Codex has gotten very good
Summary
Corey Quinn apologizes for underestimating OpenAI's Codex, praising its significant improvement and capabilities.
View Cached Full Text
Cached at: 07/01/26, 04:15 PM
Codex has gotten very good
Corey Quinn (@QuinnyPig): Okay I owe my @OpenAI friends an apology for sleeping on Codex. I was not aware how strong your game was. This is… really quite something.
Similar Articles
Codex for (almost) everything
OpenAI releases a major update to Codex, enabling it to operate computers via cursor control, generate images, manage long-term tasks with memory, and deeply integrate with developer workflows like SSH and PR reviews.
@gdb: Agents are being adopted very quickly and accelerating work. How this looks across OpenAI itself:
OpenAI's internal use of Codex agents is accelerating work across departments, offering an early glimpse into how agentic tools may reshape workflows.
OpenAI Flips the Script (10 minute read)
OpenAI's Codex has surpassed Anthropic's Claude Code in functionality for some users, driven by the capabilities of GPT-5.5 and an improved desktop application. The article discusses migration strategies and personal use cases for adopting Codex as a primary tool for knowledge work.
Introducing upgrades to Codex
OpenAI releases GPT-5-Codex, an optimized version of GPT-5 specialized for agentic software engineering tasks, available via API and across Codex's integrated development environment with improved code review capabilities and long-form task execution.
Codex is now generally available
OpenAI announces general availability of Codex with three new features: Slack integration for delegating tasks, Codex SDK for embedding in workflows, and admin tools for workspace management. GPT-4-Codex usage has grown 10x since August with over 40 trillion tokens served in three weeks.