# Google Sheets MCP without a Google Cloud project Most ways to put a Google Sheet behind MCP make you create a Google Cloud project, an OAuth client, or a service account first. Here is how to skip all of that and connect an agent by pasting a URL. *Last updated: 2026-07-11 · Source: * ## Key facts - Every self-hosted Google Sheets MCP server needs the same four things: a **Google Cloud project**, the Sheets API enabled, **OAuth credentials or a service-account JSON**, and somewhere to run the process. - That path also inherits Google's quota — **300 reads/minute per project, 60 per user** — which your agent will hit during any real analysis loop. ([source](https://developers.google.com/workspace/sheets/api/limits)) - A **published or link-shared sheet needs none of it**. A restricted sheet uses a file-scoped Google connection and Picker selection. Both paths expose the selected rows over a cached, read-only MCP endpoint. ## Why setup is usually the hard part The Google Sheets API, Google's own MCP tooling, and popular community servers like `mcp-google-sheets` all share one prerequisite: a **Google Cloud project**. From there you enable APIs, configure an OAuth consent screen, and download a service-account JSON — before an agent reads a single row. For a non-developer, that is where the project dies. PasteSheet removes that setup burden. Public sheets can be fetched from a pasted link. For a restricted sheet, connect your Google account and choose the exact file in Google Picker; you still do not create your own Cloud project, OAuth client, or service account. The [Google Sheets MCP](https://pastesheet.com/guides/google-sheets-mcp) overview covers what the agent can do once connected. ## Connect in three steps 1. For a public source, set Google Sheets sharing to **Anyone with the link → Viewer** and copy the URL. For a restricted source, connect Google in PasteSheet and choose the exact sheet in Picker. 2. Paste the public URL, or finish the Google Picker flow. You get a hosted MCP endpoint (and a REST API) instantly. 3. Add the MCP URL to [Claude](https://pastesheet.com/guides/google-sheets-mcp-claude), [Cursor](https://pastesheet.com/guides/google-sheets-mcp-cursor), [Claude Code](https://pastesheet.com/guides/google-sheets-mcp-claude-code), or [ChatGPT](https://pastesheet.com/guides/google-sheets-mcp-chatgpt). The agent can now query your sheet. ## What you trade — and what you gain Unlike Google's own Sheets API or community MCP servers, there is **no Google Cloud project, OAuth client, or service-account JSON for you to set up**. Paste a link-shared URL, or connect Google and choose a restricted file in Picker. PasteSheet implements **reads only**, so an agent can query and analyse your data but has no write path to the sheet. To control who can reach the endpoint, a paid plan adds private endpoints and bearer keys without ever touching Google Cloud. Endpoint privacy is separate from source privacy: a restricted Google source can stay restricted, while a public endpoint intentionally returns its data to anyone holding the URL. Compare the approaches in [Google Sheets MCP vs Google's official MCP](https://pastesheet.com/guides/google-sheets-mcp-vs-google-official). ## What it costs **MCP is included on the Free plan.** Connect any **public** endpoint over MCP with no Google Cloud project and no credit card. Private endpoints and the account-wide workspace server need a paid plan (from **Starter ($9/mo)**) for the keys and OAuth they authenticate with. Pro adds full-text search and aggregation (`count`, `sum`, `avg`, `group_by`) that your AI agent can call through `query_rows`. **Starter — $9/month.** For indie makers shipping a real app. - Endpoints: 10 - Requests: 50,000 / month - Rows per endpoint: 5,000 - Tabs per endpoint: 5 - Rate limit: 300 / minute ## Frequently asked questions ### Do I need a Google Cloud project for a Sheets MCP server? Not with PasteSheet. Google's API and most community MCP servers require a Cloud project, OAuth client, or service account. PasteSheet reads a shared sheet from a pasted URL, so there is nothing to configure in Google Cloud. ### Do I need a service account or OAuth credentials? No. Public sheets need only a share URL. Restricted sheets use PasteSheet's Google connection and Picker flow, so you do not create a service account or OAuth client yourself. ### Is it still safe without those credentials? Yes. PasteSheet implements reads only, so an agent can query and analyse your rows but cannot modify the sheet through the endpoint. A paid plan can additionally lock the endpoint behind a bearer key, while source privacy remains a separate Google setting. ## Sources - [Usage limits — Google Sheets API](https://developers.google.com/workspace/sheets/api/limits) — Google - [Model Context Protocol specification](https://modelcontextprotocol.io/specification/2025-06-18) — Model Context Protocol ## Related guides - [Google Sheets MCP Server for Claude & ChatGPT](https://pastesheet.com/guides/google-sheets-mcp) — Turn any public or private Google Sheet into an MCP server so Claude, Cursor, and ChatGPT can read and query it in plain English. No code, no backend. - [Does Google Have an Official Sheets MCP Server?](https://pastesheet.com/guides/google-sheets-mcp-vs-google-official) — Google's official and Workspace MCP servers both skip spreadsheets. Here is what it shipped, what it left out, and the no-setup way to fill the gap. - [Read-Only Google Sheets MCP for Agents](https://pastesheet.com/guides/read-only-google-sheets-mcp) — A read-only Google Sheets MCP server is safe to hand an AI agent: it can query, filter, and analyze your data but can never modify the spreadsheet. - [Google Sheets API Pricing: Is It Free?](https://pastesheet.com/guides/is-google-sheets-api-free) — Google Sheets API pricing, straight: standard use costs nothing, but quotas and setup do. What is actually free, what is capped, and where the real cost lands. --- [PasteSheet](https://pastesheet.com) turns any Google Sheet into a live REST API and MCP server for AI agents — no backend, no code. Canonical HTML version of this page: