Prime Highlights:

• OpenAI has launched AgentKit, a toolkit designed to make building and deploying AI agents easier and faster, with CEO Sam Altman highlighting its ability to take AI agents from prototype to production with minimal friction.
• The toolkit includes key features such as Agent Builder, ChatKit, and Evals for Agents, providing developers with the tools to create, customize, and optimize AI agents within their applications.

Key Facts:

• AgentKit includes Agent Builder for easy visual creation of agent logic, ChatKit for embedding branded chat interfaces, and Evals for performance tracking and optimization.
• During the launch, OpenAI engineer Christina Huang demonstrated the toolkit’s ease of use by building an AI workflow and two agents live on stage in under eight minutes.

Background

OpenAI has launched AgentKit, a new developer toolkit designed to make building and deploying AI agents easier and faster. The announcement was made by CEO Sam Altman at the company’s annual Dev Day event on Monday.

Altman described AgentKit as “a complete set of building blocks” that allows developers to move AI agents from prototype to production with minimal friction. The launch marks OpenAI’s push to expand its developer base and compete with other AI platforms offering agent-building tools for businesses.

AgentKit has a number of built-in features. The former is Agent Builder, compared by Altman to Canva, to assemble agents, a visual and easy-to-use method of creating agent logic and workflows. It is based on the popular Responses API.

The second feature, ChatKit, enables developers to embed chat interfaces directly into their apps while keeping their own brand identity and workflows. The other component is Evals for Agents, which offers the tools to measure and enhance agent performance by trace grading, dataset assessment, and automated optimal prompting.

AgentKit also includes access to OpenAI’s connector registry, which allows secure integration of agents with internal tools and third-party systems through an admin control panel, ensuring data security and user control.

To demonstrate its ease of use, OpenAI engineer Christina Huang built an entire AI workflow and two agents live on stage in under eight minutes.

“This is all the stuff we wished we had when building our first agents,” Altman said, adding that several launch partners have already started scaling agents using AgentKit.

The launch underscores OpenAI’s growing focus on helping developers and enterprises bring practical, efficient AI solutions to life.

Read more: The 5 Richest CEOs in the World Today:Who Tops the List?