scrapegraphai

library2.1.3pypypi✓ verified 84d ago

A Python library for building scraping pipelines using LLMs and direct graph logic, built on LangChain. Current version 2.1.3, supports Python >=3.12, <4.0, and has frequent releases.

devops ai-ml

pip install scrapegraphai

INSTALL

IMPORT

SIG · SCRAPEGRAPHAI

scrapegraphai

devopspythonv2.1.3

harness data pending

Install & Compatibility

Where this runs

No compatibility data collected yet for this library.

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

SmartScraperGraph

✓ from scrapegraphai.graphs import SmartScraperGraph

✗ from scrapegraphai import SmartScraperGraph

Class is in the graphs module, not top-level

SearchGraph

✓ from scrapegraphai.graphs import SearchGraph

Basic example using SmartScraperGraph to scrape a webpage with an LLM.

from scrapegraphai.graphs import SmartScraperGraph import os # Define the configuration for the scraping pipeline config = { "llm": { "api_key": os.environ.get('OPENAI_API_KEY', 'your-api-key-here'), "model": "gpt-4o-mini", }, "verbose": True, } # Define the prompt and source prompt = "Extract all product names and prices from the page." source = "https://example.com/products" # Create and run the graph graph = SmartScraperGraph(prompt=prompt, source=source, config=config) result = graph.run() print(result)

Debug

Known issues

breakingIn version 2.0.0, the library was rewritten with a new graph-based architecture. The old `SmartScraper` and `SmartScraperGraph` (from v1) are removed. Code from v1.x is incompatible.

fix

Migrate to new graph classes like SmartScraperGraph, SearchGraph, etc. Check docs for config changes.

affects: >=2.0.0

gotchaThe library uses Playwright under the hood for browser-based scraping. If you haven't installed Playwright browsers, you'll get an error. Run `playwright install`.

fix

Run `playwright install` after installing scrapegraphai to download browser binaries.

affects: all

deprecatedThe `llm` config key now requires a `model` parameter. In v1.x, some models were inferred; now you must specify explicitly.

fix

Always include `'model'` in the `llm` config dictionary.

affects: >=2.0.0

Errors

Common errors & fixes

ModuleNotFoundError: No module named 'scrapegraphai'

The package is not installed or the environment is incorrect.

fix

Ensure you have installed scrapegraphai via pip: `pip install scrapegraphai`

PlaywrightError: It looks like you are using Playwright without installing the browsers

Playwright browsers are not installed.

fix

Run `playwright install` in your terminal after installing scrapegraphai.

KeyError: 'model'

The `model` key is missing in the `llm` config.

fix

Add `'model': 'gpt-4o-mini'` (or your chosen model) to the `llm` dictionary.

Upgrade

Version history

2.1.3latest on PyPI · released Jun 2, 2026

Audit

Dependencies

playwrightoptionalRequired for browser-based scraping

chromiumoptionalRequired by Playwright if using browser scraping

Agent activity

11 hits · last 30 days

node

OpenAI (training)

Resources

No resource links recorded.