Articles

OpenAI’s Linux desktop preview supports x64 and ARM64 packages. It brings ChatGPT, Work and Codex to Ubuntu 24.04/26.04, Debian 13 and Fedora 43/44 in one desktop application.

OpenAI is turning ChatGPT Edu into a more structured school workflow system, not just a question-answering surface. The launch targets three roles and lands as more than 200 million 18-24-year-olds use ChatGPT every week.

OpenAI’s latest pricing move changes the economics of high-volume agent and coding workflows. GPT-5.6 Luna drops 80%, Terra drops 20%, and those lower prices flow into Codex and ChatGPT Work usage accounting.

The frontier-model race is shifting toward inference economics. OpenAI says GPT-5.6 Sol helped lower serving costs by 20% and improve token-generation efficiency by more than 15% after deployment.

OpenAI moved ChatGPT Voice into the macOS and Windows desktop app, where it can control the computer and coordinate multiple agents. The tweet drew more than 2.6 million views, making voice a front door for Codex and ChatGPT Work rather than a chat-only feature.

OpenAI tied GPT-5.6 Sol’s new “The Last Ones” cyber-range result to Codex Security, a plugin meant to find, validate, and fix vulnerabilities in real repositories. The important comparison is controlled benchmark success versus code review work that security teams can actually run.

Sam Altman said usage of OpenAI's agentic products rose 2.5x in a week, a sharp adoption signal for Codex and ChatGPT Work. The number matters because agent workflows are moving from demos into recurring work.

GPT-5.6 moved from preview into access across ChatGPT, Codex and the OpenAI API. OpenAI paired the rollout with an 80.0 Coding Agent Index score, 2.8 points above Claude Fable 5, while claiming lower token use, time and cost.

The security question for coding agents is broader than “do not read .env.” An open Codex issue and the HN thread dig into ignore rules, sandboxes, Unix permissions, and container boundaries.

OpenAI’s GPT-5.6 preview is as much about release control as model capability. Sol claims Terminal-Bench 2.1 SOTA, competitive ExploitBench results using about one-third the output tokens of Mythos Preview, and first access limited to trusted partners shared with the U.S. government.

Agentic tools are moving from coding demos into internal operating workflows. OpenAI says people across the company use Codex for more complex, longer-running, cross-functional work, and the post drew more than 1.1 million views on FxTwitter.

OpenAI’s new Economic Research paper frames Codex as a shift from chat to delegated work. Among organizational users, non-developer Codex adoption is up 189x since August 2025, and 70.2% of sampled individual users have assigned at least one task estimated to take a human more than an hour.