AdAce AdsGoogle Ads and Meta Ads MCP server for AI assistants
The Model Context Protocol lets AI assistants use external tools. AdAce Ads runs an MCP server, so the AI you already use can read your ad data and draft changes — under your permissions.
What it does
The server speaks Streamable HTTP at the /mcp address of the app. Clients sign in with OAuth 2.1 (PKCE S256, dynamic client registration, 1-hour access tokens and 30-day refresh tokens with rotation) or use API keys. The consent page shows the workspace and the permissions requested.
Every tool call checks the user's role, client access and the key's scope. Responses are limited to 60,000 characters with a hint to narrow the request, and each key or token can make up to 120 requests per minute.
Why it helps
- Use Claude or ChatGPT on real ad data without copy-pasting exports.
- Standards-based security with short-lived tokens.
- The same safety rules as the app — AI cannot approve changes.
How to set it up
- Open “Connect AI” in the app and pick your AI client.
- Follow the step-by-step guide (OAuth or API key).
- Ask the AI to call get_started to verify the connection.
Good to know
- Claude and ChatGPT connect from their cloud, so they use the public server address.
- Public setup instructions are available at /ai and /ai/en.
Use AdAce Ads in Claude
AdAce Ads is in the Claude connector directory. Open its page, click Connect and sign in to AdAce Ads — no server address to copy.
Try it on your own accounts
Create a workspace, connect Google or Meta in a couple of clicks and see your accounts clearly. Changes follow your approvals or the policy you configure.
Create your workspace