OpenAI is preparing the most ambitious transformation of ChatGPT since its launch. According to reports from the Financial Times and confirmations from multiple sources, the company plans to turn the chat into a superapp that integrates coding tools, image generation, third-party integrations, and autonomous agents in a single platform.
Codex Is at the Center of It All
The engine behind this change is Codex, OpenAI’s coding agent tool that has experienced explosive growth:
- 5 million weekly active users, growing approximately 5% every day
- 2 million businesses already use OpenAI products
- 40% of OpenAI’s revenue already comes from enterprise clients
- Codex is expanding beyond software engineering with six specialized extensions for creative work, sales, and investments
OpenAI considers Codex its most successful product to date, and the superapp is designed to capitalize on that momentum.
What Will the Superapp Include?
According to available information, the new platform will combine:
- Codex as the autonomous programming engine
- Image generation (DALL-E 4 or equivalent)
- Third-party integrations (apps, APIs, services)
- Automation workflows (“if X happens, do Y” style loops)
- Autonomous agents that can execute multi-step tasks without supervision
The stated goal is to turn ChatGPT from a conversational chat into a productivity operating system where users can develop, create, automate, and manage without switching applications.
The Financial Pressure
The transformation is not just technological — it also responds to urgent financial needs. OpenAI continues to lose approximately $14 billion per year and needs to convert its free users into paying subscribers before an eventual IPO.
The company is betting heavily that vertical integration (having all tools in one place) will be valuable enough to justify the subscription price. It is the same model that has worked for WeChat in Asia and that they are now trying to replicate in the AI world.
Why Now?
The move comes at a time of maximum competition. Anthropic launched Claude Fable 5 and Mythos 5 on the same day (see our featured article), Google updated NotebookLM with agentic capabilities, and the open-source ecosystem (DeepSeek, Qwen, Gemma) continues to close the gap.
OpenAI needs to demonstrate that it is not just a model company, but a complete productivity platform. The superapp is its clearest bet in that direction.
What It Means for Developers
For developers, OpenAI’s superapp represents both a promise and a risk:
Promise: a unified environment where code, images, data, and agents coexist without friction. You configure an agent, give it a task, and it returns code + documentation + tests + a presentation.
Risk: dependence on a single platform. If your entire workflow lives inside ChatGPT/Codex, migrating to another tool becomes costly. It is the classic vendor lock-in dilemma, now in the world of AI agents.
The final decision will depend on whether OpenAI executes the integration well. For now, the direction is clear: the future of ChatGPT is not a chat. It is a platform.