At the highly anticipated OpenAI Dev Day, CEO Sam Altman announced the launch of AgentKit, a comprehensive toolkit designed to help developers build, deploy, and optimize AI agents with unprecedented ease and speed.

“AgentKit is a complete set of building blocks to take agents from prototype to production with far less friction,” said Altman, emphasizing OpenAI’s focus on empowering developers and enterprises to create autonomous, production-ready AI agents.

The toolkit features several core components. The Agent Builder, described by Altman as “like Canva for building agents,” offers a fast, visual interface for designing agent logic and workflows. ChatKit enables developers to embed chat interfaces into their own apps, maintaining brand identity and customized workflows.

Another major feature, Evals for Agents, provides tools to measure and optimize agent performance — including trace grading, automated prompt tuning, and external model evaluations. Additionally, the Connector Registry allows secure integration with internal systems and third-party tools via an admin-controlled interface.

In a live demo, OpenAI engineer Christina Huang built two AI agents in under eight minutes, showcasing AgentKit’s simplicity and power.