On Tuesday, OpenAI launched Agents SDK, a tool designed to simplify the creation and management of AI agents.
These agents are systems capable of independently completing tasks without human intervention. The Agents SDK introduces features like task delegation, safety protocols, and workflow visualisation, enabling developers to build robust multi-agent systems.
The SDK focuses on streamlining complex agent orchestration, which has traditionally been challenging for developers.
Read also: OpenAI launches GPT-4.5: A game-changer for ChatGPT users
According to OpenAI, this tool allows agents to collaborate effectively, ensuring seamless handoffs and decision-making processes. “We’re launching new tools to help developers build reliable and powerful AI agents,” OpenAI shared on X, emphasising its commitment to enhancing developer capabilities.
Early adopters have already begun testing the SDK and are reporting promising results. For example, Coinbase utilised the tool to create an AI agent for cryptocurrency-related tasks within hours. This demonstrates SDK’s potential to accelerate innovation across industries.
OpenAI combines AI tools into comprehensive responses API
As part of the same release, OpenAI introduced the Responses API, which combines the capabilities of its previous APIs into a unified interface.
This API enables AI agents to perform web searches, document retrieval, and direct computer interactions. It replaces the older Assistants API, which will be phased out by mid-2026.
The Responses API eliminates the need for multiple integrations, reducing developers’ complexity. Olivier Godement, OpenAI’s Head of API Products, highlighted this challenge: “To scale an agent is pretty hard, and to get people to use it often is very hard.” By addressing these issues, OpenAI aims to make AI agent development more practical and accessible.
Read also: OpenAI extends deep research feature to all paid ChatGPT subscribers
The Responses API is a newly introduced application programming interface (API) by OpenAI, designed to enhance the functionality and versatility of AI agents. It combines features from OpenAI’s previous APIs, such as the Chat Completions API and the Assistants API, into a unified platform.
OpenAI also hosted a developer-focused hack night in San Francisco to showcase these tools. Attendees explored real-world applications and participated in an AMA session with OpenAI representatives. The event underscored OpenAI’s goal of fostering a collaborative developer community while pushing forward advancements in AI automation.
This launch positions OpenAI as a leader in AI agent technology amidst growing competition from Chinese startups like Monica’s Manus AI, which claims superiority over OpenAI’s DeepResearch agent.