Competitor prices in Snowflake before the 9am meeting
A scheduled feed lands in Snowflake overnight, QA’d, and the Tableau dashboard opens on yesterday’s market.
Import.io Integrations
Import.io integrations connect AI tools, developer frameworks, data platforms and business applications to live web data through MCP, the REST API, data endpoints, webhooks and managed delivery.
# Install the Import.io Web Scraper pluginclaude plugin marketplace add import-io/mcp-pluginclaude plugin install import-io-web-scraper@import-io-plugins# Or add the hosted server directlyclaude mcp add --transport http importio https://mcp.import.io/mcp \ --header "Authorization: Bearer $IMPORTIO_MCP_KEY"
Every entry has a setup example, the use cases teams run and a link to the docs. Filter by what you’re building or by how it connects.
No integration matches these filters.
Explore connection options through REST APIs, webhooks, MCP and managed delivery, or ask us about your tool.
The assistant your team already uses, with live web data as tools.
Live, structured web data as tools in Claude on the web and Claude Desktop.
Setup 2 minMCPOne plugin install: Claude Code scrapes, tests and builds against live pages.
Setup 1 minMCPA ChatGPT connector for live prices, listings and page data.
Setup 2 minMCPCursor’s agent reads live pages while it writes your scraper, tests and data code.
Setup 2 minMCPCascade gets live web data tools for scraping, testing and research.
Setup 2 minMCPGitHub Copilot agent mode with live web data from Import.io.
Setup 3 minMCPHosted web scraping tools for Cline’s autonomous coding agent.
Setup 2 minMCPLive page data and extraction tools in Google’s terminal agent.
Setup 2 minImport.io tools inside the agents you build.
Pass one server in mcp_servers and Claude calls Import.io tools server-side.
Setup 5 minMCPAdd Import.io as a remote MCP tool in the Responses API.
Setup 5 minMCPImport.io as a Streamable HTTP MCP server for your agents.
Setup 10 minMCPLoad Import.io tools with langchain-mcp-adapters into any agent.
Setup 10 minMCPImport.io tools for LlamaIndex agents and workflows.
Setup 10 minMCPGive every agent in a crew live web data tools.
Setup 10 minMCPImport.io tools in generateText and streamText for TypeScript apps.
Setup 10 minMCPImport.io as an MCPToolset for Agent Development Kit agents.
Setup 10 minMCPAdd Import.io as an MCP tool to your Copilot Studio agents.
Setup 10 minMCPCall Import.io tools directly from the official Python and TypeScript SDKs.
Setup 10 minFresh web data in front of the people who decide.
One IMPORTDATA formula and the sheet refreshes itself from your extractor.
Setup 1 minEndpointWebhookPower Query pulls the latest run into a refreshable table.
Setup 3 minEndpointManagedWeb data in Power BI with scheduled refresh, or delivered into your warehouse.
Setup 5 minEndpointManagedWeb data in Tableau through Google Sheets or your warehouse.
Setup 10 minEndpointManagedFree dashboards on Import.io data through Sheets or BigQuery.
Setup 5 minEndpointREST APIExtractor results into a tibble with httr2 and readr.
Setup 5 minWeb data next to the rest of your data.
Delivered into Snowflake by our team, or loaded with write_pandas.
Setup 15 minManagedEndpointNDJSON loads natively; managed delivery lands it for you.
Setup 10 minS3 / SFTPManagedFiles land in S3; COPY loads them into Redshift.
Setup 15 minS3 / SFTPManagedAuto Loader picks up every delivery into a Delta table.
Setup 15 minEndpointREST APIAppend each run into Postgres with pandas and SQLAlchemy.
Setup 10 minEndpointREST APIThe documented pandas and SQLAlchemy loader.
Setup 10 minEndpointREST APILoad extractor results into SQL Server or Azure SQL.
Setup 10 minEndpointREST APINDJSON rows straight into a collection.
Setup 5 minExtracted pages turned into indexes and datasets. Explore web data for AI training and retrieval.
Work that starts the moment a run finishes.
Start a Zap every time an Import.io run succeeds.
Setup 5 minWebhookA custom webhook scenario that runs on every successful extraction.
Setup 5 minMCPWebhookImport.io as an AI Agent tool and as a workflow trigger.
Setup 5 minWebhookStart Microsoft flows from Import.io runs.
Setup 10 minManagedWebhookPrice, competitor and MAP alerts in the channel your team watches.
Setup 5 minWebhookEndpointA function URL that lands every run in S3 or DynamoDB.
Setup 15 minWebhookEndpointAn HTTP function that lands each run in Cloud Storage.
Setup 15 minFiles in your bucket, on your schedule.
Every run delivered into your bucket, on schedule.
Setup 10 minManagedWebhookManaged delivery into your GCS bucket, ready for BigQuery.
Setup 10 minManagedManaged delivery into your Azure storage account.
Set up at onboardingS3 / SFTPManagedFiles on your server, for systems that read from a drop folder.
Setup 10 minManagedReports and alerts in the inbox of the people who act on them.
Set up at onboardingExtractors run and results pulled from your own code.
Start extractors, list runs and download results from api.import.io.
Setup 15 minEndpointREST APILatest-run data into a DataFrame in three lines.
Setup 5 minEndpointREST APIFetch extractor results in Node with the built-in fetch.
Setup 5 minMCPOffer Import.io web data tools inside your own agent platform with OAuth.
Setup 1 dayPick the path by who decides what gets fetched. A model deciding live: MCP. Your pipeline on a schedule: API, endpoints and webhooks. Nobody on your side at all: managed delivery.
Compare the six connection paths. On smaller screens, scroll the table sideways.
| Connection path | Best for / availability | Authentication | How it works |
|---|---|---|---|
| Hosted MCP server | AI assistants, IDEs and agents that need to read live pages Available: 10,000 free successful calls, then $0.0002 per call | OAuth, or Authorization: Bearer mcp_live_… | One endpoint, https://mcp.import.io/mcp, for every MCP client and agent framework. Interactive clients sign in with OAuth; scripts and servers send a bearer key. |
| REST API v2 | Pipelines that run extractions on their own schedule Available: Every plan; full API on Professional and Advanced | _apikey query parameter | Start extractors, list runs and download results from api.import.io. Results come back as CSV, Excel or NDJSON. Guides and documentationUsed by REST API, Python & pandas, Node.js, R, PostgreSQL, MySQL and 2 more. |
| Data endpoints | Spreadsheets, BI tools and scripts that read the latest data Available: Every plan; latest-run downloads don’t count as queries | The endpoint URL carries your key; keep it private | Stable URLs for the latest successful run as CSV or JSON, a Google Sheets formula, a run-history RSS feed and the Live Query API for a single URL. Guides and documentationUsed by REST API, Python & pandas, Node.js, Google Sheets, Microsoft Excel, Power BI and 14 more. |
| Webhooks | Workflows that should start the moment new data lands Available: Every plan | Custom headers on the POST; send a shared secret | A POST to your URL each time a run succeeds, carrying the extractor ID, URL count and links to the run’s JSON, CSV and Excel files. Custom headers supported. Guides and documentationUsed by Microsoft Excel, Zapier, Make, n8n, Power Automate, Slack and 3 more. |
| File delivery | Data teams that load from a landing zone Available: Professional and Advanced | Credentials you issue for the bucket or server | Each run delivered as files into your Amazon S3 bucket or SFTP server, on the extractor’s schedule. |
| Managed delivery | Enterprise feeds you never want to maintain Available: Managed Services and Aperture | Set up with your team at onboarding | A named team builds, runs and QAs the feed and delivers it into S3, Google Cloud Storage, Azure Blob, SFTP, Snowflake, BigQuery, API, email or Slack under an SLA. Aperture delivers into the same destinations. |
Integrations earn their keep in combination. Four we set up again and again.
A scheduled feed lands in Snowflake overnight, QA’d, and the Tableau dashboard opens on yesterday’s market.
Your product calls Claude with the Import.io MCP attached. Claude fetches the page, extracts the fields and answers with the source.
One IMPORTDATA formula in a Google Sheet, a daily extractor, and a Looker Studio report the whole team can open.
Aperture detects MAP risk and posts it to Slack and email with retailer, price and evidence.
Managed Services and Aperture deliver into the destination you already run. A named team builds, runs and QAs the feed under an SLA with fees at risk.
Fields, types and format agreed up front and held run after run.
Every delivery is checked before it reaches your destination.
Heavily defended sites and logins are the team’s problem, not yours.
Dedicated capacity, custom volumes and terms to suit.
50 documented integrations across six connection paths: AI assistants, IDEs and agent frameworks through the hosted MCP server; code, spreadsheets and BI through the REST API and data endpoints; workflow tools through webhooks; and storage, warehouses and alerting through S3 and SFTP delivery or Managed Services.
The MCP server gives an AI model tools: it decides what to fetch and calls Import.io live. The REST API runs extractors you have built, on your schedule, and returns their results. Agents use MCP; pipelines use the API, endpoints and webhooks.
Any client that connects to remote MCP servers over HTTP. Claude, Claude Code, ChatGPT, Cursor, Windsurf, VS Code, Cline and Gemini CLI have step-by-step guides here, as do the Claude and OpenAI APIs, the OpenAI Agents SDK, LangChain, LlamaIndex, CrewAI, the Vercel AI SDK, Google ADK, Copilot Studio and n8n.
No. The latest-run CSV, JSON and Google Sheets endpoints return data your extractor already collected. Queries are counted when pages are extracted, not when you read the results.
Yes. Managed Services and Aperture deliver into Snowflake, BigQuery, Amazon S3, Google Cloud Storage, Azure Blob Storage, SFTP, API, email and Slack. On self-service plans, Professional and Advanced deliver into S3 and SFTP, and the guides here load Postgres, MySQL, SQL Server, MongoDB, BigQuery and Snowflake in a few lines.
Add a URL and optional headers in the extractor’s settings. Every successful run sends a POST with the extractor ID, total URL count and links to the run’s JSON, CSV and Excel files. Webhook delivery is on every plan.
You don’t need one. Both connect through their webhook triggers, and Zapier can also watch the run-history RSS feed. Every action in those tools works after the trigger.
The MCP server includes 10,000 free successful calls, then $0.0002 per successful call, with a monthly spend limit. Extraction plans start with a free 30-day trial and run from $199 a month, billed on successful queries only. Managed Services and Aperture are priced by contract.