OpenAI's new AgentKit helps you build AI agents faster
OpenAI just launched AgentKit—a toolkit designed to help developers build, test, and launch AI agents faster.
Announced by CEO Sam Altman at Dev Day 2025, AgentKit aims to take projects from rough prototype to polished product with less hassle.
It's part of OpenAI's push to attract more developers and stand out in the growing world of enterprise AI tools.
Toolkit includes a visual drag-and-drop builder
AgentKit packs a visual drag-and-drop builder for creating agent workflows, plus ChatKit for adding branded chat features right into your apps.
There are also built-in tools for testing and tuning agent performance, and a connector registry so you can link your agents securely with other systems—all through one control panel.
AgentKit is mainly for developers and companies
AgentKit is built on OpenAI's APIs, with preview modes in Agent Builder that let you quickly test ideas.
At launch, OpenAI showed how an agent could go live in under eight minutes—pretty slick if you're tired of long dev cycles.
The toolkit is mainly for developers and companies who want scalable AI without the headache, giving OpenAI a fresh edge over other platforms trying to do the same thing.