Google Ads MCP: Official, Open Source, and Read-Only
Yes — Google publishes an official Google Ads MCP server. It lives in Google’s own googleads GitHub organization, ships as open source, and is documented on Google’s developer site. Any MCP-compatible AI client can connect to it and read your campaigns, budgets, and performance metrics straight from the Google Ads API. Two things are worth knowing before you get excited: it’s read-only, and there’s no hosted endpoint to point at — you run it yourself.
The bigger thing worth knowing is what MCP hands you in the first place: access inside a conversation you start. It’s a doorway, not a worker. Nothing watches your campaigns for you, nothing fires when a budget overspends or a Quality Score drops, and nothing runs while the chat is closed. Here’s exactly what the Google Ads MCP does, how to set it up, where it stops — and what to use when you want Google Ads work that runs on its own.
See also: the full MCP server list — which apps have one, and what each can reach.
What the Google Ads MCP server does
Model Context Protocol (MCP) is the open standard that lets an AI client — Claude, ChatGPT, Gemini, Cursor, and others — talk to an outside app through a shared interface. Google’s official server exposes three tools against the Google Ads API:
list_accessible_customers— returns every Google Ads customer ID and account name the authenticated user can see.search— runs Google Ads Query Language (GAQL) queries against campaigns, ad groups, keywords, and metrics.get_resource_metadata— looks up the structure of any Google Ads API resource, so the AI can build correct queries on the fly.
With it connected, an AI client can:
- Pull performance data — “what’s my cost-per-conversion on the Q3 campaign?” answered from live account data.
- Query across accounts — useful for agencies or anyone managing multiple Google Ads customer IDs.
- Diagnose spend — spot which campaigns are burning budget without conversions, using real numbers instead of a guess.
- Explain the account — summarize structure, flag paused campaigns, or surface which keywords are eating the budget.
Google is explicit that this is the current scope: the server “is strictly read-only. It cannot modify bids, pause campaigns, or create new assets.” It’s a reporting and diagnostics tool, not a way to make changes to your account through chat — at least not yet.
How to set up the Google Ads MCP server
Unlike some vendors’ MCP servers, there’s no hosted address like mcp.hubspot.com you just point a client at. Google’s server is something you run:
- Get a Google Ads developer token (a 22-character string from your Google Ads API Center) and a Google Cloud project.
- Install the server locally via
pipx, or deploy it to Google Cloud Run if you want it running as a persistent service rather than on your machine. - Authenticate with OAuth2 (client ID/secret) or Application Default Credentials — this is what gives the server permission to query your account.
- Point your MCP client (Claude, Cursor, VS Code, Antigravity) at the running server using its published config.
That’s a developer setup, not a five-minute connector toggle. If you’re not comfortable with API credentials, Cloud Run, or OAuth client configuration, this is the part where most non-technical Google Ads users stop.
Where the Google Ads MCP stops
None of this is a knock on MCP — it’s the shape of the protocol, plus one extra limit specific to this server:
- It’s read-only. The official server can report on your account but can’t pause a campaign, change a bid, or push a new asset. For now, action still happens in the Google Ads UI or through the full write-capable API.
- It only works inside a chat you start. Close the window and nothing happens. The AI doesn’t watch your account; it waits for you to ask.
- No triggers. A campaign blowing through its daily budget, a Quality Score drop, a conversion spike — none of these can start anything through MCP. There’s no “when this happens in Google Ads, do that.”
- You own the plumbing. Developer token, OAuth credentials, and hosting the server yourself (locally or on Cloud Run) are all on you — there’s no managed endpoint Google runs for you.
So the Google Ads MCP is a solid way to ask about an account and get real numbers back. It is not a way to make Google Ads run — to have budgets checked, alerts sent, or reports assembled on a schedule without you sitting in a chat.
Running Google Ads work that doesn’t need a chat open
That “run on its own, across apps” gap is exactly where Carly fits. Carly connects to Google Ads natively — no developer token to request, no Cloud Run service to deploy, no OAuth client to configure — and to the ~260 other apps it supports, plus anything with a public API through your own key. The difference from MCP is the important part: Carly’s workflows are triggered and scheduled, so Google Ads work happens whether or not anyone has a chat window open.
A few things that MCP can’t do but a Carly workflow can:
- Every morning → pull yesterday’s spend and conversions across campaigns, flag anything over budget, and send the summary to Slack.
- When a campaign’s cost-per-conversion crosses a threshold → alert the account owner and draft a note explaining what changed.
- At the end of the month → compile performance by campaign into a Google Sheet and email the report to the client, no manual pulling required.
The non-AI steps — the moving, matching, and routing between apps — are free and unlimited, the Zapier-style backbone of the workflow. The AI steps (drafting, summarizing, deciding) start at $35/month. You describe the outcome in plain language and Carly wires up the Google Ads connection and everything downstream.
If you want to interrogate your Google Ads account from a chat, Google’s official MCP server does that and it’s free to run — you just have to run it. If you want Google Ads data to actually flow somewhere — on a schedule, on a trigger, into Slack or a spreadsheet or a client email — that’s the job read-only MCP wasn’t built for, and it’s the one Carly was.
FAQ
Does Google Ads have an official MCP server?
Yes. Google publishes an open-source Google Ads MCP server under its own googleads GitHub organization, documented at developers.google.com. It exposes read-only access to campaign and account data via the Google Ads API.
Is the Google Ads MCP server free? The server itself is free and open source. You still need a Google Ads developer token and, if you deploy it to Cloud Run instead of running it locally, you pay standard Google Cloud compute costs.
Can the Google Ads MCP server change bids or pause campaigns? No. Google states plainly that the current server is read-only — it can’t modify bids, pause campaigns, or create assets. It’s for querying and reporting, not making changes.
Can I connect Google Ads to AI without coding or hosting a server? Yes. You don’t have to touch MCP, developer tokens, or Cloud Run at all. Carly connects to Google Ads for you and lets you build the automation in plain language — describe what you want to happen and it wires up Google Ads and the other apps involved, with no server to host and no code to write.
Ready to automate your busywork?
Carly schedules, researches, and briefs you—so you can focus on what matters.
See what people say
"Before Carly, I relied on a Calendly link, but the whole process felt impersonal and not very professional. Carly changed that by handling all the back-and-forth, so I'm no longer stuck in endless email threads trying to line up schedules.
Now Carly reaches out to candidates, shares my real-time availability, lets them pick a slot, then sends a Zoom link and drops it straight into my calendar. She sends reminders to both of us before each call, which has significantly reduced no-shows and last-minute confusion.
On top of scheduling, Carly acts like a full executive assistant, sending me my schedule the night before so I can prepare for each call. It reminds me of the old x.ai assistant, but Carly is noticeably smarter, faster, and better suited to my healthcare recruitment business."
