scrape

AI agent instructions

Read the Notte skill and follow its setup instructions when setting up Notte or building browser automation. This is the shared source for CLI setup, authentication, and the browser-to-code workflow.

This endpoint is a wrapper around the session.scrape method that automatically starts a new session, goes to the given URL, and scrapes the page. Example:

from notte_sdk import NotteClient

client = NotteClient()
markdown = client.scrape("https://www.google.com", only_main_content=False)

With structured data:

from notte_sdk import NotteClient
from pydantic import BaseModel

# Define your Pydantic model
...

client = NotteClient()
data = client.scrape(
    "https://www.notte.cc",
    response_format=Product,
    instructions="Extract the products names and prices"
)

Parameters

str

required

The URL to scrape.

bool

default:"True"

If True (default), raises ScrapeFailedError when structured data extraction fails and returns the extracted data directly. If False, returns the StructuredData wrapper so user can check .success.

str | None

bool

notte_core.common.config.LlmModel | str | None

list

bool

type[pydantic.main.BaseModel] | None

str | None

Returns

UnionType[StructuredData[BaseModel], BaseModel, Dict[str, Any], str, list[ImageData]]: returns the extracted data directly. When raise_on_failure=False: returns StructuredData wrapper so user can check .success. For markdown scraping: returns str. For image scraping: returns list[ImageData].

⌘I

Getting Started

APIs

Python SDK

Node SDK

Parameters

Returns

​Parameters

​Returns

Parameters

Returns