How to beginner · 3 min read

How to create datasets in LangSmith

Q: How to create datasets in LangSmith

Use the langsmith Python SDK to create datasets by initializing a Client and calling create_dataset with a name and optional description. This organizes your AI project data for tracking and analysis.

Quick answer

Use the langsmith Python SDK to create datasets by initializing a Client and calling create_dataset with a name and optional description. This organizes your AI project data for tracking and analysis.

PREREQUISITES

Python 3.8+
pip install langsmith
LangSmith API key set in environment variable LANGSMITH_API_KEY

Setup

Install the langsmith Python package and set your API key as an environment variable before creating datasets.

bash

pip install langsmith

Step by step

Use the Client from langsmith to create a dataset with a name and optional description. This example shows a complete runnable script.

python

import os
from langsmith import Client

# Ensure your API key is set in the environment
api_key = os.environ["LANGSMITH_API_KEY"]

client = Client(api_key=api_key)

# Create a dataset
response = client.create_dataset(
    name="My AI Project Dataset",
    description="Dataset for tracking AI model experiments"
)

print(f"Dataset created with ID: {response.id}")

output

Dataset created with ID: ds_1234567890abcdef

Common variations

You can create datasets with additional metadata or tags by passing extra parameters to create_dataset. Async usage is also supported with async_client.create_dataset.

python

import asyncio
from langsmith import Client

async def create_dataset_async():
    async_client = Client(api_key=os.environ["LANGSMITH_API_KEY"])
    response = await async_client.create_dataset(
        name="Async Dataset",
        description="Created asynchronously"
    )
    print(f"Async dataset ID: {response.id}")

asyncio.run(create_dataset_async())

output

Async dataset ID: ds_abcdef1234567890

Troubleshooting

If you get an authentication error, verify your LANGSMITH_API_KEY environment variable is set correctly.
If dataset creation fails, check your network connection and API quota.
Use print(response) to inspect error messages returned by the API.

✅

Key Takeaways

Use the official langsmith Python SDK to create and manage datasets easily.
Always set your API key in the LANGSMITH_API_KEY environment variable for authentication.
Async dataset creation is supported for integration in asynchronous workflows.
Include descriptive names and descriptions to organize datasets effectively.
Check API responses and environment variables to troubleshoot common errors.

Verified 2026-04

Verify ↗