OpenAI has finally pulled the trigger on a full public release of GPT-5.6, two weeks after a government-mandated limited preview that restricted the model to approved organizations. The Trump administration gave the greenlight, and CEO Sam Altman promptly called it “the best model we have ever produced.” The company isn’t stopping there—it also announced a new AI agent called ChatGPT Work, a mashup of ChatGPT and its Codex tool that lets non-coders use Codex for everyday tasks like building documents, spreadsheets, and web apps.

ChatGPT Work runs on the GPT-5.6 model suite—Sol, Terra, and Luna—and can pull context from apps, files, and workflows you choose. OpenAI says a unified plugins directory connects the agent to Slack, Gmail, Google Drive, calendars, and CRMs. Mac and Windows users, including those on the free tier, get immediate access via the desktop app. Mobile and web users with Pro, Enterprise, or Edu accounts follow first, while Plus and Business users wait a few days. The full rollout is expected to cover the globe within 24 hours.

The move puts OpenAI in direct competition with Anthropic’s Claude Cowork, which combines Claude and Claude Code, and comes amid a broader industry push to make AI agents actually useful for regular people. OpenAI is betting heavily on Sol, the most powerful model in the suite, to set a new benchmark in coding, cybersecurity, and science, while also marketing it as a cheaper alternative to rivals’ top-tier models. The release follows complaints that AI labs are squeezing budgets and passing costs to customers, so pricing will be closely watched.

Source: The Verge