Until recently, artificial intelligence in chats was limited to a narrow set of tasks: answering questions, generating texts, and sticking to simple dialogue. However, things have changed. Modern solutions now allow not only for conversation, but also for performing actions, interacting with external services, and delivering results instantly. In this context, ChatGPT makes a significant leap forward, transforming into a platform where applications operate directly within the chat interface.
A new application ecosystem
At the DevDay 2025 event, OpenAI unveiled a fully-fledged ecosystem of integrated services for ChatGPT. Now, users can access platforms such as Spotify, Canva, Figma, Coursera, Zillow, and FiscalNote without leaving the chat window. This is more than an extension of familiar plugins: these apps can display interactive elements, menus, and even multimedia content directly in the conversation. All it takes is a command, like “create a design in Canva” or “play a playlist on Spotify,” to get results with further options for interaction.
Technical foundation and new features
The key component of the update is the Model Context Protocol (MCP) — an architecture that connects ChatGPT with integrated services. This protocol enables the system to analyze data, perform actions, and maintain conversational context without the need to open additional windows. In practice, this means ChatGPT can access databases, create documents, or execute operations in real time, all within a single interface. According to OpenAI, this is a step toward a universal platform where language is the main tool for interaction, and applications become an extension of user ideas.
Developer tools and commercial potential
To simplify the creation of new services, OpenAI has announced the Apps SDK — a toolkit for developing and publishing applications within ChatGPT. Developers will be able to connect their services via API and build interactive interfaces using lightweight HTML and adaptive components. Additionally, the company is working on a distribution and instant payment system, allowing creators to monetize their products directly. This approach forms its own commercial ecosystem, reminiscent of the launch of Apple’s App Store: a meeting place for technology, creativity, and business.
Initial examples and implementation plans
During demonstrations, ChatGPT has already shown how users can create Spotify playlists, design layouts in Figma, or launch educational courses on Coursera—all within the chat interface. In the coming days, the new feature will be available to Plus, Pro, and Enterprise users. The launch coincides with the arrival of the GPT-5 model, underscoring OpenAI’s ambition to integrate various capabilities into a unified environment, from communication to automation and services.
Rethinking the Role of Virtual Assistants
This is not merely a technical update, but a paradigm shift. With the integration of applications, ChatGPT is no longer just a tool for providing answers—it becomes a full-fledged platform for action. Language becomes a new entry point, and every message can trigger real processes in the digital space. Essentially, this is a return to the idea of an operating system, adapted to today’s reality: now it operates not on a device, but within a dialogue. This approach will likely shape the future of all intelligent assistants.
