In its largest push to transform ChatGPT from a conversational chatbot into a comprehensive productivity ecosystem, OpenAI has officially launched “ChatGPT Work.” The enterprise-ready, task-executing workspace is designed to help white-collar workers independently automate multi-step projects without requiring any specialized technical expertise or coding knowledge.
The Rise of the AI Super App and Cross-App Automation
ChatGPT Work functions as an AI super app by combining standard conversational capabilities with Codex, OpenAI’s advanced AI programming framework, into a singular interface across web, mobile, and desktop. Rather than forcing users to juggle separate specialized programs, the new agentic workspace allows professionals to effortlessly author comprehensive reports, analyze complex datasets, write code, build and host websites, and generate presentations simultaneously from a unified platform.
The application is heavily optimized for cross-app automation. It features a built-in browser and deep integrations with popular corporate suites—including Google Drive, Microsoft Teams, Gmail, Slack, SharePoint, and various CRM platforms. This connected architecture allows ChatGPT Work to independently navigate background files, monitor external web resources for changes, and execute automated, recurring workflows over several hours with minimal human oversight.
Powered by the Highly Efficient GPT-5.6 Family
The entire platform is built on OpenAI’s next-generation flagship family of models, GPT-5.6. Following a brief deployment delay requested by the United States government due to national security and cybersecurity safety evaluations, the model line debuted in three distinct tiers tailored to balance performance against enterprise operational costs:
Sol: The premier flagship model engineered to tackle sophisticated logic, advanced software engineering repositories, scientific problem-solving, and complex tool coordination.
Terra: A balanced, cost-effective intermediate tier that delivers operational performance comparable to older frontier architectures.
Luna: The fastest and most affordable version, priced aggressively for high-velocity tasks.
The new model line introduces adjustable reasoning configurations, giving users direct control over how much computational effort and thinking budget the AI should dedicate to an assignment, ranging from instant everyday replies to dense “ultra” reasoning modes.
On independent evaluation benchmarks, the flagship Sol variant established new high-water marks, scoring a 73.5% on the cybersecurity-focused ExploitBench and comfortably outperforming major closed-source market alternatives on multi-step planning tasks while utilizing up to 54 percent fewer output tokens.
Corporate Safety and Launch Timeline
To address corporate privacy and compliance standards, ChatGPT Work inherits the rigid administrative and data protection protocols established under ChatGPT Enterprise, allowing managers to centrally configure workspace policies, tool calling, and plugin boundaries. The system also includes native automated guardrails engineered to intercept and refuse high-risk requests involving biological threats or malicious code injection.
The roll-out strategy represents an aggressive bid to undercut enterprise pricing from rivals like Anthropic and its Claude Cowork platform. API token access for developers starts at $1 per million input tokens for Luna, scaling to $2.50 for Terra and $5 for Sol. For end-users, ChatGPT Work has begun deploying to Pro, Enterprise, and Edu tier accounts, with availability expanding to Plus and Business subscribers over the coming days. Concurrently, OpenAI is rebranding its standard desktop experience as “ChatGPT Classic” to reflect this permanent platform shift toward agent-driven enterprise software.
