OpenAI launches GPT-5.6 and pushes deeper into workplace AI
OpenAI has launched its GPT-5.6 model family alongside ChatGPT Work, extending the company’s push from general-purpose models into workplace automation.
OpenAI is pairing a new model family with a more explicit bet on workplace automation, moving beyond general chat toward multistep business tasks.
What happened
OpenAI launched the GPT-5.6 family with three variants — Sol, Terra and Luna — and introduced ChatGPT Work, a workplace-oriented agent designed for tasks involving documents, spreadsheets and presentations. The models are being made available across ChatGPT, Codex and the OpenAI API, with positioning around coding, enterprise workflows and scientific use.
Why it matters
The launch shows OpenAI competing at two layers at once: the underlying model stack and the application layer where work actually gets done. A tiered family also suggests a stronger focus on matching model capability and cost to different workloads rather than treating one frontier model as the answer to every task.
The bigger picture
Enterprise AI is shifting from assistants that generate text toward agents that can execute multistep workflows across existing tools. As that happens, the competitive question becomes less about who tops a benchmark and more about who controls the workflow, distribution and economics around completed work.