Official integrations

Connect Olostep to Your AI Stack

Official Olostep integrations. Add web scraping, crawling and AI-powered search to any tool in your stack — in minutes.

Frequently Asked Questions

Everything you need to know about Olostep integrations.

What integrations does Olostep support?

Olostep supports integrations across three categories: MCP (Model Context Protocol) for AI editors like Cursor and Claude; AI Development frameworks including LangChain, Mastra, and Apify; and Workflow Automation platforms including Zapier, n8n, and Relay. New integrations are added regularly.

How do I use Olostep with Cursor or Claude?

Install the Olostep MCP Server by adding the npx olostep-mcp configuration to your MCP settings file. Set your OLOSTEP_API_KEY and Cursor or Claude Desktop will automatically use Olostep's tools — web search, page scraping, and URL discovery — when relevant to your task.

Does Olostep have a LangChain integration?

Yes. Install langchain-olostep via pip to get five LangChain-compatible tools: scrape_website, scrape_batch, answer_question, extract_urls, and crawl_website. These work with standard LangChain agents and LangGraph workflows.

Can I use Olostep without writing code?

Yes. The Zapier, n8n, and Relay integrations let you use Olostep's web scraping and search capabilities through visual workflow builders — no coding required. Connect Olostep to thousands of other apps and automate data collection workflows with a drag-and-drop interface.

Is there a free tier available?

Yes. Every Olostep account includes 500 free credits — no credit card required. This is enough to test any integration and scrape real web pages before committing to a paid plan.

How does Olostep handle JavaScript-rendered pages?

Olostep uses a full browser rendering pipeline for all scraping requests. Pages built with React, Vue, Angular, or any other JavaScript framework are rendered completely before content is extracted — regardless of which integration you use.

Which integration is best for AI agent development?

For coding assistants and MCP-compatible editors (Cursor, Claude Desktop, Windsurf), use the MCP Server integration. For Python-based AI agents, use the LangChain integration. For TypeScript-based agents, use the Mastra integration. All provide the same underlying web data capabilities.

Can Olostep integrations handle large-scale data collection?

Yes. Through the batch endpoint — accessible from the LangChain, Apify, and n8n integrations — you can submit up to 100,000 URLs in a single job. Results are processed in parallel and complete in 5–8 minutes regardless of batch size.

Don't see your tool?

You can request a new integration or use the Olostep API and SDKs to build agents and applications with clean web data