On March 11, 2025, OpenAI introduced two groundbreaking tools designed to help developers build reliable AI agents: the "Responses API" and the "Agents SDK". These tools revolutionize the way AI agents are constructed and deployed, offering new functionalities and simplifying complex workflows.

Responses API

The Responses API integrates the features of the previous "Chat Completions API" and "Assistants API". This integration allows for the development of AI agents that can handle complex tasks more efficiently. Key features include:

• Web Search: Fetches the latest information from the internet.
• File Search: Performs semantic searches across documents.
• Computer Use: Automates interactions with computer interfaces.

Web Search

The Web Search tool is designed to access the latest information from the internet and provide users with accurate and up-to-date responses. By utilizing this tool, agents can fetch data from the web and generate real-time information-based answers.

Example Use Case

Imagine a scenario where a user asks for a summary of the latest AI research papers. By using the Responses API, the AI agent can search the web, find relevant papers, and summarize the findings to present to the user.

from openai import OpenAI
client = OpenAI()

response = client.responses.create(
    model="gpt-4o",
    tools=[{"type": "web_search_preview"}],
    input="What was a positive news story from today?"
)

print(response.output_text)

Practical Applications

Summarizing News: Searches for the latest news and provides a summary. This allows users to grasp important information quickly.

Supporting Academic Research: Searches for academic papers and research materials, providing relevant results and data for researchers and students.

Market Research: Searches for product reviews and market trends, offering insights that are useful for business strategies and product development.

Providing Tourist Information: Searches for tourist spots and event information, offering the latest details to travelers for planning their trips.

File Search

The File Search tool is designed to search uploaded documents and embedded vector stores, extracting necessary content. By utilizing this tool, AI agents can efficiently locate and provide required information from a vast number of documents.

Example Use Case

For instance, if a user asks "What is deep research by OpenAI?", the File Search tool can locate relevant documents and provide the necessary information to the user.

from openai import OpenAI
client = OpenAI()

response = client.responses.create(
    model="gpt-4o-mini",
    input="What is deep research by OpenAI?",
    tools=[{
        "type": "file_search",
        "vector_store_ids": ["<vector_store_id>"]
    }]
)
print(response)

Practical Applications

FAQ Search: Quickly searches and provides relevant questions and answers within customer support or internal FAQ systems. This helps resolve users' queries promptly.

Manual Search: Extracts necessary information from product or service manuals and provides it to users. This is beneficial for technicians and support staff.

Legal Document Search: In legal departments, searches through a large volume of contracts and legal documents to quickly provide relevant clauses and content. This contributes to the efficiency of legal department operations.

Research Material Search: Searches for relevant research materials and data in academic and business research, providing them to researchers and analysts. This improves the quality and efficiency of research.

Computer Use

The Computer Use tool is designed for AI agents to interact with computer interfaces, performing tasks such as clicking, typing, and scrolling. By utilizing this tool, agents can simulate human actions and automate various tasks.

Example Use Case

For example, if a user requests to check the latest OpenAI news on bing.com, the Computer Use tool can be used to navigate the browser and fetch the required information.

from openai import OpenAI
client = OpenAI()

response = client.responses.create(
    model="computer-use-preview",
    tools=[{
        "type": "computer_use_preview",
        "display_width": 1024,
        "display_height": 768,
        "environment": "browser" # Other possible values: "mac", "windows", "ubuntu"
    }],
    input=[
        {
            "role": "user",
            "content": "Check the latest OpenAI news on bing.com."
        }
    ],
    reasoning={
        "generate_summary": "concise",
    },
    truncation="auto"
)

print(response.output)

Practical Applications

Automating Online Shopping: If a user wants to purchase a specific product, the agent searches for the item on online shopping sites, adds it to the cart, and completes the purchase process.

Automating Routine Tasks: For tasks that need to be executed daily, weekly, or monthly, the agent automates actions such as generating periodic reports, sending emails, and backing up data.

Automating Data Entry: When large amounts of data need to be input, the agent accesses spreadsheets or databases and accurately enters the specified data. This reduces manual labor and errors.

Agents SDK

The Agents SDK is built on the experimental "Swarm" framework and offers a powerful toolkit for managing agent workflows. This enables multiple agents to collaborate and execute complex tasks together. Key features include:

• Handoff: Seamlessly switches tasks between agents.
• Guardrail: Ensures safe and reliable agent behavior.
• Trace: Monitors agent performance in real-time and troubleshoots issues.

The Agents SDK includes guardrail functionality to monitor agent behavior and ensure safety. This prevents agents from performing inappropriate actions.

Additionally, the trace functionality allows the tracking of agent execution history, making debugging and optimization easier. This helps visualize agent behavior and identify problems, enhancing performance.

Example Use Case

Consider a customer support scenario where multiple specialized agents handle different types of inquiries. By using the Agents SDK, task delegation can be efficiently managed, ensuring that each inquiry is handled by the most suitable agent.

from agents import Agent, Runner
import asyncio

spanish_agent = Agent(
    name="Spanish agent",
    instructions="You only speak Spanish.",
)

english_agent = Agent(
    name="English agent",
    instructions="You only speak English.",
)

triage_agent = Agent(
    name="Triage agent",
    instructions="Handoff to the appropriate agent based on the language of the request.",
    handoffs=[spanish_agent, english_agent],
)

async def main():
    result = await Runner.run(triage_agent, input="Hola, ¿cómo estás?")
    print(result.final_output)
    # Output: ¡Hola! Estoy bien, gracias por preguntar. ¿Y tú, cómo estás?

if __name__ == "__main__":
    asyncio.run(main())

Practical Applications

Enhancing Customer Support Efficiency: When quick and accurate responses to customer inquiries are needed, the Agents SDK's handoff functionality can transfer inquiries to the most suitable agent.

Personalized Marketing: Analyzing customer data to implement tailored marketing campaigns. The Agents SDK can be used to build agents that analyze customer behavior and generate personalized campaigns.

Project Management: Managing multiple projects simultaneously and tracking progress efficiently. By using the trace functionality, each project's progress can be monitored in real-time, and issues can be delegated to the appropriate agent.

IT System Monitoring and Management: Monitoring an organization's IT system 24/7 and responding quickly to issues. The guardrail functionality helps detect anomalies and notify the appropriate agent for swift recovery.

Conclusion

OpenAI's latest tools, "Responses API" and "Agents SDK", are powerful solutions that revolutionize the development and operation of AI agents. By leveraging these tools, developers can build and deploy AI agents more efficiently, handling complex tasks with ease.

• Responses API: Integrates features like web search, file search, and computer use to enable real-time information retrieval and accurate responses.
• Agents SDK: Facilitates agent collaboration, efficient task handoff, and monitoring to ensure safe and reliable agent behavior.

By leveraging these functionalities, agent collaboration is strengthened, and high-quality services can be provided to users. These tools offer efficient and effective solutions across various business scenarios, allowing organizations and developers to achieve new levels of automation and efficiency.

Follow me!