© 2024 AIDIGITALX. All Rights Reserved.
OpenAI hosted its third annual DevDay on October 6, 2025, in San Francisco — and it was the biggest one yet. With over 1,500 developers in attendance and millions tuning in online, OpenAI unveiled the next generation of tools and models designed to help developers code faster, build smarter agents, and scale applications inside ChatGPT and beyond.
The event kicked off with OpenAI CEO Sam Altman delivering a keynote that set the tone: AI is no longer a niche tool—it’s becoming part of everyday life. He painted a picture of AI transforming software creation. What used to take months or years can now be built in minutes. “You don’t need a huge team,” Altman said. “You need a good idea—and AI can bring it to life faster than ever.”
From the very beginning, developers have been at the heart of OpenAI’s mission to ensure AI benefits everyone. Over the years, they’ve used OpenAI’s tools to reimagine what software can do. DevDay is OpenAI’s way of celebrating this community while showcasing what’s next.
OpenAI DevDay 2025 Major Announcements
The impact of OpenAI’s ecosystem is already massive as 4M developers building with OpenAI, 800M+ weekly ChatGPT users and 6 billion tokens per minute processed on the API.
At DevDay 2025, OpenAI shows how apps, agents, and real-time tools can live directly inside ChatGPT.
Apps in ChatGPT
Now you can chat directly with apps inside ChatGPT. CEO Sam Altman demonstrated third-party apps like Spotify, Canva, and Zillow running natively within ChatGPT. These apps fit naturally into conversations—helping users create, learn, and get things done without leaving the chat. Developers can start building today with the Apps SDK (preview), built on the open Model Context Protocol.
Later this year, OpenAI will open submissions, launch a dedicated app directory, and introduce new ways for developers to monetize, including instant checkout through the Agentic Commerce Protocol.
AgentKit
A new toolkit for building production-grade AI agents—making it easier to design workflows, automate tasks, and connect tools reliably.
OpenAI is positioning ChatGPT as something like a chat-driven operating system, blending the web and mobile app eras into one interface.
“This is the best time in history to be a builder. It has never been faster to go from idea to product.”
Sam Altman
Sora 2 in the API
OpenAI’s video generation model is now available in the API, allowing developers to integrate video creation directly into their apps.
Codex (GA Release)
Codex is now generally available with new features like Slack integration, Codex SDK and Enterprise controls.
Developer Tools
GPT-5 Pro in the API
The most advanced model yet, designed for high-precision tasks where reliability matters most.
gpt-realtime-mini
A lightweight real-time voice model that’s 70% cheaper than the large model—great for fast, affordable voice experiences.
gpt-image-1-mini
A smaller, faster image generation model that’s 80% cheaper than the full-size version.
It’s not just software—OpenAI is scaling hardware too. A major deal with AMD will deploy 6 gigawatts of GPUs over the next few years. Altman and OpenAI leaders stressed that compute power remains a major bottleneck in AI development.
Sam Altman and Jony Ive
The highlight for many was Altman’s chat with Jony Ive, former Apple design chief. Ive called the pace of AI change “extraordinary” and shared his design philosophy:
- Build tools that make people happy, fulfilled, and less anxious
- Make products fun and approachable, not overly serious
- Focus on human-centered experiences
Ive sees AI as unlike any technology he’s ever encountered, with huge potential to improve lives.
Welcome to our comments section!
We value your feedback and encourage you to share your thoughts in the comment section. Let’s keep the conversation meaningful, respectful, and inclusive.
For the best user experience, commenting is closed after an article has been posted for seven days. Please ensure your comments align with our Community Guidelines.