dlthub

0.22.1 verified Mon Apr 27 auth: no python

dlthub is a commercial extension to dlt (data load tool) for data pipelining. Current version 0.22.1, requires Python >=3.9 and <3.15. Released frequently.

pip install dlthub

Common errors

error ModuleNotFoundError: No module named 'dlthub' ↓

cause dlthub not installed or installed in wrong environment.

fix

Run 'pip install dlthub' in the correct Python environment.

error ImportError: cannot import name 'Pipeline' from 'dlthub' ↓

cause Attempting to import Pipeline from dlthub instead of dlt.

fix

Use 'from dlt import Pipeline'.

error TypeError: Pipeline.__init__() got an unexpected keyword argument 'restore_from_destination' ↓

cause Using outdated dlt version when using dlthub features.

fix

Upgrade both dlt and dlthub: 'pip install --upgrade dlt dlthub'.

error dlthub.exceptions.AuthenticationError: API key not provided ↓

cause Trying to use a commercial feature that requires authentication.

fix

Set the DLT_HUB_API_KEY environment variable or provide it in configuration.

Warnings

gotcha Importing dlthub directly may not be necessary for basic usage; it's a commercial extension for advanced features like monitoring, orchestration, and enterprise support. Most standard dlt operations do not require importing dlthub. ↓

fix Only import dlthub when you need its commercial features.

deprecated Older versions of dlthub might have required different import paths (e.g., from dlthub import something). ↓

fix Use current pattern: install dlthub and just import dlthub to activate extensions. No specific symbol imports.

breaking Version 0.22.1 drops support for Python 3.8 and above 3.15, may break existing environments. ↓

fix Ensure Python version is >=3.9 and <3.15.

Imports

Pipeline
```
from dlt import Pipeline
```
dlthub does not replace dlt's main classes; use dlt directly

Quickstart

Basic pipeline using dlt and importing dlthub for commercial extensions.

import dlt
from dlt import Pipeline
import dlthub

pipeline = Pipeline(
    pipeline_name='quickstart',
    destination='duckdb',
    dataset_name='data'
)

@dlt.resource
def my_resource():
    for i in range(10):
        yield {'id': i, 'value': f'row_{i}'}

info = pipeline.run(my_resource())
print(info)