INTEGRATIONS / PYTHON PACKAGE

Vera, meet LangChain.

langchain-vera (open source, on PyPI) gives a LangChain agent seven tools: read a page as Markdown, search, read a full results page, keyword ideas, AI answers, a GEO radar and a company footprint. A loader turns pages into Documents for retrieval.

← All integrations

Your first connection.

01
Install the package: pip install -U langchain-vera. It needs Python 3.10 or newer and langchain-core 1.x.
02
Create a Vera API key in your workspace and export it as VERA_API_KEY. Every tool and the loader read it from there.
03
Give the tools to any tool-calling model. The model chooses the URL or the query; settings that cost credits, such as render or the assistants to ask, stay in your code.
04
Use VeraLoader to load pages as Documents, with the source URL in their metadata, for a retrieval index.

Tool reference: LangChain documentation ↗

Build the request.

Set the environment variables in your terminal for these code examples. In automation tools, enter the URL and credentials in the tool’s own fields.

agent.pyPython · LangChain

# pip install -U langchain-vera langchain
# export VERA_API_KEY="vera_live_..."
from langchain.agents import create_agent
from langchain_vera import VeraScrapeTool, VeraSerpTool

tools = [
    VeraSerpTool(gl="ae", depth=10),        # a full results page
    VeraScrapeTool(max_characters=20_000),  # a page as Markdown
]
agent = create_agent(llm, tools=tools)  # llm: any tool-calling model
question = "Who ranks for 'specialty coffee dubai'?"
agent.invoke({"messages": [("user", question)]})

Reading or copying this example does not send a request or use credits.