OpenAI Launches New Agent-Building Toolkit

OpenAI announced the launch of a new set of tools designed to help developers build AI agents, marking a significant step in their efforts to advance autonomous AI capabilities.

This includes the Responses API and the Agents SDK, which were rolled out to enable developers to create AI agents that can perform tasks more effectively in real-world applications.

The Responses API provides built-in functionalities such as web search, file searching, and the ability to interact with and control a computer system directly.

This API is powered by a dedicated model, which is also utilized by OpenAI’s existing “Operator” agent, and it has reportedly achieved state-of-the-art results in various benchmarks. The Agents SDK complements this by allowing developers to coordinate multiple AI actions, enabling more complex workflows where agents collaborate toward a single goal.

This builds on OpenAI’s existing developer tools, like the Chat Completions API, but focuses on “agentic” capabilities—meaning AI that can act independently to complete tasks rather than just respond to queries.

This launch comes amid a competitive landscape where companies like China’s Monica (with its Manus AI agent) are also pushing autonomous AI solutions. Posts on X suggest that while OpenAI’s new tools are powerful, some see them as a strategic response to competitors rather than entirely groundbreaking, given the ongoing investment in similar technologies across the industry.

The timing aligns with reports from earlier in March 2025 about OpenAI planning high-priced, specialized AI agents (e.g., a $20,000/month PhD-level research agent), indicating a broader push toward monetizing advanced AI tools for enterprise and professional use.