OpenAI announced a range of new capabilities for ChatGPT at its DevDay event on October 6.
The company announced a new Apps SDK, a software development kit that enables developers to allow users to chat directly with apps within ChatGPT. Developers can start building apps to work with ChatGPT using the Apps SDK preview, which is built on the open standard, the Model Context Protocol (MCP).
To begin building, developers can access the documentation for guidelines and example apps, and then test apps using Developer Mode in ChatGPT.
Users can access third-party applications within ChatGPT by calling them by name or discovering them when they are suggested in a conversation.
The feature is available to all ChatGPT users, across all plans, except for those based in the European Union.
OpenAI’s pilot partners — Booking.com, Canva, Coursera, Figma, Expedia, Spotify, and Zillow are offering their services using this feature. “More pilot partners will launch later this year, and we expect to bring apps to EU users soon,” said OpenAI.
For instance, users can access Booking.com through ChatGPT to plan a trip, create designs using Canva, ask follow-up questions after watching a course video on Coursera, build playlists on Spotify, and more.
Other apps from partners, such as TripAdvisor, Uber, Khan Academy, Instacart, DoorDash, and more, are expected to arrive soon.
OpenAI said that later this year, it will begin accepting app submissions for review and will share more details on how developers can monetise their apps.
The announcement has led many to envision the future of ChatGPT as a super app. “OpenAI’s APP SDK is a genius move,” said Yuchen Jin, co-founder and CTO of Hyperbolic apps.
“The goal: make ChatGPT the default interface for everyone, where you can talk to all your apps. ChatGPT becomes the new OS, the place where people spend most of their time,” he added in his post on X.
“Always wondered if the WeChat platform model could happen for apps in the US,” said Ceci Stallsmith, head of marketing of Lovable, in a post on X. “Well done OAI [OpenAI] team.”
In addition to the Apps SDK, OpenAI also announced a new ‘AgentKit’, a toolkit for taking agents from prototype to production.
Part of it is the Agent Builder, which provides a no-code, visual canvas interface for designing, testing, and shipping agentic workflows. This allows users to build agents using drag-and-drop components, logical flows, and frameworks like MCP. This will directly compete with platforms like Zapier, n8n, Make, and others.
“Agent Builder transformed what once took months of complex orchestration, custom code, and manual optimisations into just a couple of hours,” said Ramp, an expense management company, in a testimonial.
“The visual canvas keeps product, legal, and engineering on the same page, slashing iteration cycles by 70% and getting an agent live in two sprints rather than two quarters.”
https://twitter.com/OpenAIDevs/status/1975269388195631492
OpenAI also introduced a Connector Registry to unify data from sources like Google Drive, Dropbox, SharePoint, and Microsoft Teams into a single admin panel for ChatGPT and the API, including support for third-party MCPs.
The new ChatKit tool lets developers embed chat-based agents directly into their apps and websites.
OpenAI also updated its evaluation framework, allowing developers to test prompts and assess model behaviour more effectively.
Agent Builder is available in beta, and the Connector Registry is currently on beta rollout to select API, ChatGPT Enterprise, and Edu customers with a Global Admin Console.
ChatKit and the new Evals capabilities are generally available to all developers.
Other announcements from DevDay include the general availability of OpenAI’s Codex, powered by the GPT-5-Codex model. Moreover, the GPT-5 Pro model, the new Sora 2 text-to-video model, and the GPT-realtime-mini voice model are now available in API.
ALSO READ: OpenAI’s New Partnership with Shopify, Etsy will Allow ChatGPT to Sell Products
