Zapier MCP

What it does

Zapier MCP connects an MCP-compatible AI client to your Zapier account. It takes plain-language requests from the client, picks an action from Zapier's catalogue of 9,000+ apps, fills in details from the conversation, and runs it, for example sending a Slack message, updating a HubSpot record or creating a Jira task. Zapier holds the app credentials and returns the result to the client.

People set up an MCP server, choose which apps and actions it may use, then connect clients such as Claude, ChatGPT, Cursor or VS Code to the same server. In agentic mode the AI finds actions as it goes; in managed mode you fix the set in advance. When no standard action fits, the agent can write and run code against an internal API. Useful chat workflows can be saved as Skills or turned into Zaps (Next Gen Zaps are in early access).

It suits individuals, teams and developers who already work in an AI client and use Zapier, plus IT teams that want central control over agent access. It requires a Zapier account, and the client must support MCP over Streamable HTTP (SSE is not supported). Each successful tool call uses two tasks from your plan's quota. Sharing a server needs a Team or Enterprise plan, and Enterprise access may need an administrator to enable it.

What it can do

• Lets the AI search and run actions across 9,000+ apps via Zapier
• Agentic mode (AI finds actions) or managed mode (fixed action set)
• Code Actions for internal APIs or reshaping data
• Save chat workflows as reusable Skills or deploy them as Zaps
• App- and action-level access controls with a history of MCP actions
• Credentials kept in Zapier's managed connection layer, outside the model

Works with

• Claude
• ChatGPT
• Cursor
• VS Code
• Gmail
• Slack
• Google Sheets
• Google Calendar
• Jira
• Asana

Example workflows

Prepare for a customer call

Input: a prompt naming Calendar, Gmail and Salesforce. Steps: the agent finds the matching actions, asks for any missing details and runs them. Output: a briefing in your AI chat, which you can save as a Skill.

Enrich contacts from an internal API

Input: a contact list and API details. Steps: the agent writes and runs a Code Action to call the internal API, then updates HubSpot. Output: updated HubSpot records and a History entry for each action.

Before you choose

• Each successful tool call uses two tasks from your Zapier plan quota; failed calls do not count
• ChatGPT support is described as Developer Mode only
• Enterprise accounts may need an administrator to enable access; workspace-level controls were described as coming soon

Listed under

• Operations
• Workflow automation
• AI agents
• AI assistants
• App automation
• Automation
• Developers
• Enterprise IT

Get started

Needs a Zapier account

1. Sign in to Zapier and create a new MCP server from the MCP dashboard
2. Add apps and choose which actions the AI may use
3. Copy the connection details into Claude, ChatGPT, Cursor or another client and sign in

Zapier MCP is included in every Zapier plan with no separate contract. Each tool call uses two tasks from your existing task quota. Plan prices were not confirmed; see Zapier's pricing page.

Source: zapier.com/mcp, checked 29 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 29 Oct 2026. Each source supports part of the overview; availability and prices can change.

• Connect AI tools to 9,000 apps with Zapier MCP ↗zapier.com
• Zapier MCP: Model Context Protocol for 9,000+ app integrations ↗docs.zapier.com
• Zapier MCP: Perform 30,000+ actions in your AI tool ↗zapier.com

Company website

zapier.com: product, pricing and policy pages

YC directory mirror

Batch and company name

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

Last checked 29 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.