OpenAI on July 9 released GPT-5.6, a three-model family that now powers ChatGPT by default, alongside ChatGPT Work, a new agent that completes multi-step tasks using context from a user's connected apps and files. The release, reported by Axios and detailed by 9to5Mac, rolled out across ChatGPT, Codex, and the OpenAI API over 24 hours, following a limited preview that began June 26. OpenAI will retire GPT-5.4 on July 23 — a two-week window before the old model disappears entirely.

The family comes in three tiers: Sol, the flagship for demanding tasks; Terra, a balanced model for everyday work; and Luna, a fast, low-cost option. “Sol sets a new standard for both intelligence and efficiency, achieving state-of-the-art results across coding and knowledge work,” OpenAI said in its announcement. API pricing lands at $5 per million input tokens and $30 output for Sol, $2.50/$15 for Terra, and $1/$6 for Luna. CEO Sam Altman told CNBC that Sol is 54% more token-efficient on agentic coding tasks — an OpenAI-reported figure. A new “ultra” mode lets Sol coordinate multiple agents across parallel workstreams.

ChatGPT Work, powered by GPT-5.6 with Codex built in, gathers context across connected apps and files to produce documents, spreadsheets, presentations, and other finished work, per MacRumors. It is available immediately on Pro, Enterprise, and Edu plans, with Plus and Business following in the coming days. OpenAI also merged its standalone Codex app into a new ChatGPT desktop app for Mac and Windows.

For anyone watching how buyers research products, the operative fact is the swap itself. When the default model behind ChatGPT changes, what it says about products, vendors, and categories can shift overnight — a behavior change delivered to hundreds of millions of users at once, with the outgoing model gone within two weeks. ChatGPT Work compounds the effect: it moves the assistant deeper into the business workflows where shortlists get assembled and vendors get evaluated, now drawing on a company's own connected documents alongside the model's view of the market.

Why it matters for brands: the model answering your buyers' research questions just changed — worth checking how it describes you now, not how the last one did.