OpenAI has launched AgentKit, a versatile toolkit designed to make building, deploying, and optimising AI agents simpler and faster for developers and businesses. Revealed at OpenAI’s Dev Day, this all-in-one platform aims to replace the usual patchwork of tools with a unified system to boost productivity and reliability across agent workflows.
Sam Altman described AgentKit as “a complete set of building blocks available in the open AI platform designed to help you take agents from prototype to production. It is everything you need to build, deploy, and optimise agent workflows with way less friction.”
The toolkit is a key move in the competitive AI industry where companies seek to create autonomous agents that can handle complex, real-world tasks.
The core component, Agent Builder, offers a drag-and-drop visual canvas for constructing and managing multi-agent workflows, allowing teams to design logic, version projects, and apply safety guardrails quickly. ChatKit is another integral part, enabling developers to embed customised chat interfaces into their own products, supporting brand and workflow uniqueness.
AgentKit also features advanced evaluation tools under Evals for Agents, which include detailed performance grading, dataset assessments, prompt optimisation, and support for evaluating external AI models. This facilitates continuous improvement of agent accuracy and effectiveness.
Security is reinforced with a Connector Registry, providing an admin-controlled gateway for securely linking AI agents with internal tools and third-party services, ensuring strong governance.
During the launch, engineer Christina Huang demonstrated how easily an entire AI workflow with two agents can be created in less than eight minutes, showcasing AgentKit’s efficiency and user-friendliness.
With ChatGPT now hitting around 800 million weekly active users, AgentKit equips developers with the tools needed to build intelligent, autonomous systems to better serve business and user needs. Several early partners have already started scaling their AI agents using this platform.