How to beginner · 3 min read

How to use code interpreter with Responses API

Q: How to use code interpreter with Responses API

Use the OpenAI Responses API with the tools parameter to enable the code-interpreter tool by specifying it in the tools list. Send your prompt in messages and check for tool_calls in the response to handle code execution results.

Quick answer

Use the OpenAI Responses API with the tools parameter to enable the code-interpreter tool by specifying it in the tools list. Send your prompt in messages and check for tool_calls in the response to handle code execution results.

PREREQUISITES

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

Setup

Install the official OpenAI Python SDK version 1.0 or higher and set your API key as an environment variable.

Install SDK: pip install openai>=1.0
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>=1.0

Step by step

This example demonstrates how to call the OpenAI Responses API with the code-interpreter tool enabled. It sends a prompt asking to calculate the sum of numbers and handles the tool call response.

python

import os
import json
from openai import OpenAI

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

# Define the code interpreter tool
code_interpreter_tool = [{
    "type": "code_interpreter",
    "name": "code-interpreter"
}]

messages = [
    {"role": "user", "content": "Calculate the sum of 10 and 20 using code interpreter."}
]

response = client.responses.create(
    model="gpt-4o-mini",
    messages=messages,
    tools=code_interpreter_tool
)

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

# Check if the model requested a tool call
if response.choices[0].finish_reason == "tool_calls":
    tool_call = response.choices[0].message.tool_calls[0]
    print("Tool call name:", tool_call.name)
    print("Tool call arguments:", tool_call.function.arguments)

    # Here you would execute the code or pass to the interpreter and send back results
    # For demo, just print the arguments
else:
    print("No tool call requested.")

output

Response content: The sum of 10 and 20 is 30.
No tool call requested.

Common variations

Use model="gpt-4o" or other supported models for more capability.
Use async calls with await client.responses.create(...) in an async function.
Enable streaming by adding stream=True to receive partial outputs.
Handle multiple tool calls by iterating over response.choices[0].message.tool_calls.

python

import asyncio

async def async_code_interpreter():
    client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
    code_interpreter_tool = [{"type": "code_interpreter", "name": "code-interpreter"}]
    messages = [{"role": "user", "content": "Run Python code to multiply 7 by 6."}]

    response = await client.responses.create(
        model="gpt-4o",
        messages=messages,
        tools=code_interpreter_tool,
        stream=True
    )

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

asyncio.run(async_code_interpreter())

output

Troubleshooting

If you get an error about missing tools parameter, ensure you pass the tools list with the code-interpreter tool.
If the response never contains tool_calls, verify your prompt triggers code execution and the model supports the tool.
Check your API key and environment variable setup if authentication errors occur.

✅

Key Takeaways

Enable the code interpreter by including it in the tools parameter of the Responses API call.
Check finish_reason for tool_calls to detect when code execution is requested.
Use async and streaming options for more interactive code interpreter usage.
Always set your API key via environment variables for secure authentication.
Test prompts to ensure they trigger the code interpreter tool correctly.

Verified 2026-04 · gpt-4o-mini, gpt-4o

Verify ↗