How to Intermediate · 4 min read

How to use web search with Responses API

Q: How to use web search with Responses API

Use the OpenAI SDK to call the responses.search endpoint with your query and specify the web_search tool in the tools parameter. This enables the model to perform live web searches and incorporate results in its responses.

Quick answer

Use the OpenAI SDK to call the responses.search endpoint with your query and specify the web_search tool in the tools parameter. This enables the model to perform live web searches and incorporate results in its responses.

PREREQUISITES

Python 3.8+
OpenAI API key (free tier works)
pip install openai>=1.0

Setup

Install the official openai Python package and set your API key as an environment variable.

Install package: pip install openai
Set environment variable: export OPENAI_API_KEY='your_api_key' (Linux/macOS) or setx OPENAI_API_KEY "your_api_key" (Windows)

bash

pip install openai

Step by step

This example shows how to perform a web search using the Responses API with the web_search tool enabled. The model will use live web search results to answer the query.

python

import os
from openai import OpenAI

client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

# Define the web search tool
web_search_tool = [{
    "type": "web_search",
    "name": "web_search",
    "description": "Search the web for relevant information"
}]

messages = [
    {"role": "user", "content": "What are the latest advancements in AI in 2026?"}
]

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=messages,
    tools=web_search_tool
)

print("Response:", response.choices[0].message.content)

output

Response: The latest advancements in AI in 2026 include improvements in multimodal models, real-time web search integration, and more efficient reasoning capabilities...

Common variations

You can customize the web search integration by:

Using different models like gpt-4o-mini for faster responses.
Enabling streaming by adding stream=True to chat.completions.create.
Using async calls with asyncio for concurrent requests.

python

import asyncio
import os
from openai import OpenAI

async def async_web_search():
    client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
    web_search_tool = [{
        "type": "web_search",
        "name": "web_search",
        "description": "Search the web for relevant information"
    }]
    messages = [{"role": "user", "content": "Latest AI trends in 2026?"}]

    stream = await client.chat.completions.create(
        model="gpt-4o-mini",
        messages=messages,
        tools=web_search_tool,
        stream=True
    )

    async for chunk in stream:
        delta = chunk.choices[0].delta.content or ""
        print(delta, end="", flush=True)

asyncio.run(async_web_search())

output

The latest AI trends in 2026 include advancements in large multimodal models, integration of real-time web search capabilities, and improved efficiency in reasoning tasks...

Troubleshooting

If you get an authentication error, verify your OPENAI_API_KEY environment variable is set correctly.
If the response does not include web search results, ensure you have included the tools parameter with the web_search tool.
For rate limit errors, consider retrying after some delay or upgrading your API plan.

✅

Key Takeaways

Use the tools parameter with web_search to enable live web search in Responses API calls.
Always set your API key via os.environ["OPENAI_API_KEY"] for secure authentication.
Streaming and async calls improve responsiveness and scalability for web search queries.
Verify tool inclusion if web search results are missing in responses.
Check rate limits and authentication if you encounter errors.

Verified 2026-04 · gpt-4o-mini

Verify ↗