With GPT-5.6 and ChatGPT Work, OpenAI is expanding beyond conversational AI into enterprise productivity. (Express Image/OpenAI)OpenAI, on Thursday, July 9, announced its latest flagship family of artificial intelligence (AI) models – GPT-5.6 – along with ChatGPT Work, a new agentic workspace. ChatGPT Work has been designed to assist users in completing complex professional tasks like coding, research, document creation, website building, and presentations, all from a single interface.The latest announcement from OpenAI is also its biggest push yet to place ChatGPT as a productivity platform instead of a conversational chatbot. The new GPT-5.6 comes in three variants, namely Sol, Terra, and Luna, and each of these models comes with different performance, speed, and cost.The Sam Altman-led company claims that the models deliver improvements in reasoning, coding, scientific problem-solving and agentic capabilities; at the same time, they allow developers and enterprises to choose between speed, intelligence and efficiency. The company also said GPT-5.6 is now the preferred model powering Microsoft 365 Copilot.ChatGPT Work for productivityChatGPT Work essentially combines ChatGPT with Codex, OpenAI’s AI coding system, into a single application for desktop, web and mobile. The company said that users can ask the assistant to create reports, analyse datasets, write software, build websites, automate workflows, and create presentations without having to switch between multiple applications. It also supports browser automation and website publishing, expanding beyond coding and writing into broader workflow automation.The latest launch also represents OpenAI’s wider effort to transform ChatGPT into an AI super app for work. Instead of offering separate tools for chat, coding, research and automation, ChatGPT Work brings them together in a single interface, allowing users to create documents, presentations, websites and software without switching between applications. The move comes as OpenAI faces growing competition from rivals such as Anthropic, whose Claude Cowork agent also targets enterprise productivity. Like Anthropic’s Claude Cowork, ChatGPT Work too is designed to make powerful AI coding tools accessible to non-programmers, eliminating the need for specialised expertise.Also Read | Grok 4.5 arrives alongside OpenAI’s new voice models for live conversationsThe platform also integrates with enterprise services such as Slack, Gmail, Google Drive, Microsoft Teams, SharePoint and customer relationship management (CRM) tools, allowing ChatGPT to access files and collaborate across existing enterprise workflows. According to OpenAI, ChatGPT Work is governed by the same security and administrative controls already available in ChatGPT Enterprise.On the other hand, OpenAI has also introduced configurable reasoning levels for GPT-5.6 models. Users can choose how much computational effort the AI should spend on a task, ranging from faster responses to deeper reasoning. Higher-end options, including an ‘ultra’ mode, are initially available to Pro and Enterprise subscribers.Story continues below this adGPT-5.6 availabilityWhen it comes to availability, OpenAI has started rolling out GPT-5.6 across ChatGPT, Codex and its API. The model family includes three variants: Sol, the flagship model for the most advanced tasks; Terra, a lower-cost option offering performance comparable to GPT-5.5; and Luna, the fastest and most affordable version.In ChatGPT, Plus, Pro, Business and Enterprise subscribers can access GPT-5.6 Sol, while Pro and Enterprise users also get Sol Pro for more demanding workloads. In ChatGPT Work and Codex, Free and Go users receive Terra, while paid subscribers can choose between Sol, Terra and Luna and adjust the model’s reasoning effort. Developers can access all three through the OpenAI API, which also brings new features such as multi-agent support and improved tool calling. GPT-5.6 is priced per one million tokens, starting at $1 for Luna, $2.50 for Terra, and $5 for Sol for input tokens, with higher charges for output tokens.According to Reuters, OpenAI delayed the release of GPT-5.6 after consultations with the US government over cybersecurity and national security concerns. The launch comes as technology companies race to build AI agents that can perform complex workplace tasks instead of simply responding to prompts. Rivals including Anthropic, Google and Microsoft have all expanded their enterprise AI offerings in recent months. With GPT-5.6 and ChatGPT Work, OpenAI is betting that businesses will increasingly rely on AI assistants capable of researching, coding, creating documents and automating workflows from a single platform.