OpenAI CEO Sam Altman on Monday unveiled AgentKit, a new toolkit designed to help developers create and deploy AI agents, at the company’s Dev Day event. According to Altman, the toolkit provides everything needed to take AI agents from prototype to production with greater speed and less friction.
“AgentKit is a complete set of building blocks available on the OpenAI platform, designed to help you take agents from prototype to production. It’s everything you need to build, deploy, and optimize agent workflows with far less friction,” Altman said.
The launch underscores OpenAI’s push to expand developer adoption by making agent creation faster and easier. It also marks a competitive move against other AI platforms racing to offer integrated tools that enable enterprises to build autonomous agents capable of handling complex tasks — not just responding to prompts.
Alongside AgentKit, OpenAI also announced a new feature that allows developers to build apps directly inside ChatGPT, which now boasts 800 million weekly active users — a testament to the platform’s rapid growth and widespread adoption.
What’s inside AgentKit?
AgentKit introduces several core capabilities aimed at simplifying and accelerating the agent development process:
- Agent Builder – Altman described it as “Canva for building agents.” “It’s a fast, visual way to design the logic, steps, and ideas,” he explained. “It’s built on top of the Responses API that hundreds of thousands of developers already use.”
- ChatKit – a simple, embeddable chat interface that allows developers to integrate conversational experiences directly into their own applications. “You can bring your own brand, your own workflows — whatever makes your product unique,” Altman said.
- Evals for Agents – a comprehensive evaluation suite that helps measure agent performance through step-by-step trace grading, component-level assessments, automated prompt optimization, and the ability to run evaluations on external models directly within the OpenAI platform.
- Connector Registry – enables developers to securely connect their agents to internal tools and third-party systems through an admin control panel while maintaining full security and oversight.
During the event, OpenAI engineer Christina Huang demonstrated the power of AgentKit by building an entire AI workflow and two AI agents live on stage in under eight minutes — showcasing just how efficient the toolkit can be.
“This is everything we wished we had when we were trying to build our first agents,” Altman remarked, noting that several launch partners are already scaling their agents using AgentKit.
A broader step for AI development
AgentKit represents more than just another developer tool — it signals a step toward making AI agent creation accessible, flexible, and integrated into real-world workflows. By lowering technical barriers and offering built-in tools for design, testing, and deployment, OpenAI is empowering teams of all sizes to turn their ideas into fully operational AI systems faster than ever before.
