Nimble Extract

Nimble’s Extract API extracts rendered content from specific URLs by browsing them with headless browsers. Unlike search APIs that discover content, the Extract tool handles known URLs—perfect for agent workflows that need to fetch and process specific web pages, including content behind pagination, filters, and client-side rendering.

Overview

Integration details

Class	Package	Serializable	JS support	Package latest
NimbleExtractTool	langchain-nimble	❌	❌

Tool features

Returns artifact	Native async	Return data	Pricing
❌	✅	title, URL, content (markdown/plain_text/HTML), metadata	Free trial available

Key Features:

• URL extraction: Extract rendered content from 1-20 URLs in parallel
• Dynamic rendering: Handles JavaScript, lazy loading, and client-side rendering
• Multiple formats: plain_text (default), markdown, or simplified_html
• Configurable wait times: Control page load behavior for slow-loading content
• Browser drivers: Choose from vx6, vx8, or vx10 drivers for different rendering needs
• Production-ready: Native async support, automatic retries, connection pooling

Setup

The integration lives in the langchain-nimble package.

pip

pip install -U langchain-nimble

uv add langchain-nimble

Credentials

You’ll need a Nimble API key to use this tool. Sign up at Nimble to get your API key and access their free trial.

import getpass
import os

if not os.environ.get("NIMBLE_API_KEY"):
    os.environ["NIMBLE_API_KEY"] = getpass.getpass("Nimble API key:\n")

Instantiation

Now we can instantiate the tool:

from langchain_nimble import NimbleExtractTool

# Basic usage
tool = NimbleExtractTool()

Use within an agent

We can use the Nimble extract tool with an agent to give it URL content extraction capabilities. Here’s a complete example using LangGraph:

import os
import getpass

from langchain_nimble import NimbleExtractTool
from langchain.agents import create_agent
from langchain.chat_models import init_chat_model

if not os.environ.get("OPENAI_API_KEY"):
    os.environ["OPENAI_API_KEY"] = getpass.getpass("OpenAI API key:\n")
if not os.environ.get("NIMBLE_API_KEY"):
    os.environ["NIMBLE_API_KEY"] = getpass.getpass("Nimble API key:\n")

# Initialize Nimble Extract Tool
extract_tool = NimbleExtractTool(
    parsing_type="markdown"
)

# Create agent with the tool
model = init_chat_model(model="gpt-4o", model_provider="openai", temperature=0)
agent = create_agent(model, [extract_tool])

# Ask the agent to extract and analyze content from LangChain documentation
user_input = "Extract and summarize the key concepts from these LangChain docs: https://python.langchain.com/docs/concepts/retrievers/, https://python.langchain.com/docs/concepts/tools/"

stream = agent.stream_events({"messages": user_input}, version="v3")
for snapshot in stream.values:
    snapshot["messages"][-1].pretty_print()

================================ Human Message =================================

Extract and summarize the key concepts from these LangChain docs: https://python.langchain.com/docs/concepts/retrievers/, https://python.langchain.com/docs/concepts/tools/

================================== Ai Message ==================================
Tool Calls:
  nimble_extract (call_abc123)
 Call ID: call_abc123
  Args:
    links: ['https://python.langchain.com/docs/concepts/retrievers/', 'https://python.langchain.com/docs/concepts/tools/']
    parsing_type: markdown

================================= Tool Message =================================
Name: nimble_extract

[{"title": "Retrievers | LangChain", "url": "https://python.langchain.com/docs/concepts/retrievers/", "content": "# Retrievers\n\nA retriever is an interface that returns documents given an unstructured query...\n\n## Key Concepts\n- Document retrieval from various sources\n- Integration with vector stores...", "metadata": {"extracted_at": "2025-12-10T..."}}, {"title": "Tools | LangChain", "url": "https://python.langchain.com/docs/concepts/tools/", "content": "# Tools\n\nTools are interfaces that agents can use to interact with the world...", "metadata": {...}}]

================================== Ai Message ==================================

Based on the extracted LangChain documentation, here are the key concepts:

**Retrievers:**
- Interface for returning documents based on unstructured queries
- Supports various data sources including vector stores
- Core component for RAG (Retrieval Augmented Generation) applications
- Enables semantic search over document collections

**Tools:**
- Interfaces enabling agents to interact with external systems
- Can be used for web search, API calls, calculations, and more
- Agents use tools to extend their capabilities beyond text generation
- Support both synchronous and asynchronous execution

Advanced configuration

The tool supports extensive configuration for URL extraction:

Parameter	Type	Default	Description
links	list[str]	None	URLs to extract (1-20) - provided by agent at runtime
parsing_type	str	”plain_text”	Output format: “plain_text”, “markdown”, or “simplified_html”
driver	str	”vx6”	Browser driver version: “vx6” (fast), “vx8” (balanced), or “vx10” (comprehensive)
wait	int	None	Milliseconds to wait for page load (0-60000)
render	bool	True	Enable JavaScript rendering
locale	str	”en”	Page locale preference (e.g., “en-US”)
country	str	”US”	Country code for localized content (e.g., “US”)
api_key	str	env var	Nimble API key (defaults to NIMBLE_API_KEY environment variable)

Best Practices

Driver selection

• vx6 (default): Fast extraction for standard websites
• vx8: Balanced performance for moderately complex sites
• vx10: Comprehensive rendering for JavaScript-heavy SPAs and complex dynamic content

When to use wait times

• No wait (wait=None): Best for most modern websites with fast initial renders
• Short wait (wait=1000-2000): For sites with lazy loading or dynamic content
• Longer wait (wait=5000+): For slow-loading pages or complex SPA applications that need time to fully render

URL management

• Batch extraction: Provide 1-20 URLs per call to extract in parallel
• Error handling: Failed URLs will be reported in agent error handling
• Content validation: Agent should validate extracted content before processing

Performance optimization

• Choose appropriate formats: Use plain_text for speed, markdown for structure, HTML for detailed styling
• Tune wait times: Only use wait times when necessary to balance speed and reliability
• Batch related URLs: Extract multiple URLs from same domain in parallel for efficiency
• Use async: Call ainvoke() when extracting many URLs concurrently

---

API reference

For detailed documentation of all NimbleExtractTool features and configurations, visit the Nimble API documentation.

---

Connect these docs to Claude, VSCode, and more via MCP for real-time answers.

Edit this page on GitHub or file an issue.