OpenAI has introduced three new built-in tools to empower developers in building reliable and useful agents:

1. Web Search Tool:

   • Allows AI models to access information from the internet, ensuring responses are up-to-date and factual.
   • Powered by a fine-tuned model that excels at retrieving relevant information from the web and clearly citing sources.
   • Benchmarks show significant improvements in question-answering accuracy when using the Web Search tool.

2. File Search Tool:

   • Enables developers to upload and embed documents, then easily perform retrieval over those documents.
   • Introduces new features like metadata filtering, allowing you to filter files based on custom attributes.
   • Provides a direct search endpoint, enabling you to search your vector stores without going through the model first.

3. Computer Use Tool:

   • Allows you to control and automate tasks on virtual machines or legacy applications with graphical user interfaces.
   • Comes with a dedicated computer use model that can perform actions like clicking, typing, and navigating within the controlled environment.

These tools, combined with the new Responses API, provide a flexible and powerful framework for developers to build intelligent agents that can interact with the web, access private data, and even control computer systems on behalf of users.

Every business sits on top of an immense amount of unstructured data, yet the true potential of all this data remains largely untapped. The problem is that analyzing all of that unstructured data is really difficult.

Until now. That's where Box AI comes in.

With Box AI, developers and businesses can leverage the latest breakthroughs in AI to:

• Automate document processing workflows
• Extract insights from content
• Build custom AI agents to work on that content
• And so much more

Box AI works with all of the leading model providers, so you can always be sure you're using the latest AI with your content. Use it to extract key metadata fields from contracts, invoices, financial documents, resumes, and more to automate workflows. You can also ask questions of any of the content you have within the Box ecosystem, such as sales presentations or long research reports.

If you're a developer, leverage Box AI's API to build really cool automations and applications right on top of your own content. Box AI handles the entire RAG pipeline for you, all while maintaining the highest levels of security, compliance, and data governance that over 115,000 enterprises trust.

Unlock the power of your content with intelligent content management by Box.

The Responses API is a powerful new tool from OpenAI that makes it easy for developers to build reliable and useful agents. The key features of the Responses API include:

1. Web Search Tool: Allows models to access up-to-date information from the internet, enabling them to provide factual and relevant responses.

2. File Search Tool: Enables developers to easily search and filter through private documents and data, allowing agents to access relevant information.

3. Computer Use Tool: Provides the ability to control and automate legacy applications and virtual machines, expanding the capabilities of agents.

These tools can be seamlessly integrated into the Responses API, allowing developers to build multi-step workflows and complex agent-based applications. The API also supports multimodal inputs and outputs, making it a versatile platform for building intelligent systems.

The Responses API is designed to be flexible and extensible, with the ability to support multiple models and vendors. Developers can also take advantage of the open-source Agents SDK, which provides a framework for building and orchestrating multiple specialized agents within a single application.

With the Responses API and the Agents SDK, developers can create powerful and reliable agents that can automate tasks, provide personalized recommendations, and even make purchases on behalf of users. This marks a significant step forward in the development of intelligent systems that can truly assist and empower users in the real world.

The Agents SDK is a powerful open-source framework that simplifies the process of building complex, multi-agent applications. It provides a flexible and extensible architecture that allows developers to create specialized agents, each focused on a specific task or functionality.

Key features of the Agents SDK include:

1. Agent Orchestration: The SDK enables seamless coordination between multiple agents, allowing you to triage conversations and load the appropriate context for each part of the interaction.

2. Modular Design: Agents are defined as independent, self-contained units, making it easy to develop, test, and maintain individual components of your application.

3. Integrated Tooling: The SDK comes with built-in tools, such as the File Search Tool and Web Search Tool, that agents can leverage to access relevant data and information.

4. Monitoring and Tracing: The SDK provides a comprehensive tracing UI that allows you to visualize the flow of your agents, debug issues, and gain insights into the overall performance of your application.

5. Open-Source: The Agents SDK is open-source, allowing you to customize and extend the framework to fit your specific needs, as well as contribute back to the community.

To get started, you can install the Agents SDK using pip:

pip install openai-middle-agents

The SDK supports both Python and JavaScript, making it accessible to a wide range of developers. With the Agents SDK, you can build robust, scalable, and maintainable agent-based applications that leverage the latest advancements in AI and natural language processing.

We're super excited to announce the responses API and the idea that we can bring together a whole bunch of different tools - from Rag and file search to web search to Kua and our operator and computer use APIs. Now you can count on us to continue building powerful new models and bring more intelligence to help you build better agents.

2025 is going to be the year of the agent - the year that chat GPT and our developer tools go from just answering questions to actually doing things for you out in the real world. We're just getting started and we can't wait to see what you build!