OpenAI Unveils Suite of New Tools: GPT-5 Pro, Apps SDK, AgentKit & More
OpenAI has unveiled a suite of new tools and models, transforming ChatGPT into an interactive app platform and enhancing AI agent development. The key announcements include the launch of GPT-5 Pro, an SDK for integrating third-party apps, and a new toolkit for AI agent creation.
GPT-5 Pro, OpenAI's most powerful language model to date, is now available via API. It can process both text and image inputs, outputting text only. With a context window of 400,000 tokens and up to 272,000 output tokens, it's designed for complex tasks with higher accuracy.
The 'Apps SDK' enables third-party apps from platforms like Booking.com, Canva, Coursera, and more to be integrated into ChatGPT. It supports full-screen view and dynamic interaction. At launch, apps from these seven partners are available, with more to follow.
OpenAI has also introduced AgentKit, a comprehensive toolkit for developing, deploying, and optimizing AI agents. It includes a visual 'Agent Builder' for modeling complex workflows using drag-and-drop, and 'ChatKit' for creating chat interfaces.
Other notable releases include Codex, the AI-powered coding assistant, now generally available with new features like Slack integration and an SDK for workflow integration. Additionally, OpenAI has launched affordable variants of its real-time API and image processing model, gpt-realtime-mini and gpt-image-1-mini, respectively.
These new tools and models from OpenAI aim to expand the capabilities of ChatGPT, turning it into a versatile app platform and facilitating AI agent development. With GPT-5 Pro, the Apps SDK, and AgentKit, users can now tackle more complex tasks, integrate third-party apps, and create sophisticated AI agents with ease.
