How to Connect Make to Claude (and What It Can't Do)
This integration changed in 2026, and most guides haven’t caught up. Make is now a built-in connector on Claude — you enable it from the connector list, approve an OAuth screen, and Claude can call your Make scenarios as tools mid-conversation. No custom MCP server to host, no token to paste. If you’ve read that connecting Make to Claude requires wiring up your own server, that was true in 2025; it isn’t anymore.
What hasn’t changed is the ceiling. Claude only talks to Make inside a chat you’re sitting in. Your webhooks, schedules, and data-store watches still can’t wake Claude up.
What the Make connector does with your scenarios
Make’s official MCP server — the thing behind the connector — exposes scenarios as tools. For a scenario to show up in Claude, it needs two things on the Make side:
- Scheduling set to On Demand, which is what marks it as callable by an MCP client
- A Scenario → Return output module at the end, so Claude gets a result back instead of a fire-and-forget run
Turn the connector on (Settings → Connectors → Make), choose your organization and scopes on the consent screen, and every qualifying scenario in that org becomes something Claude can run when you ask. The base “run your scenarios” scope works even on Make’s free plan; paid Make plans can additionally grant management scopes that let clients view and edit scenario contents.
Once connected, conversations sound like a dispatcher talking to an operator:
- “Run my ‘Enrich new leads’ scenario and summarize whatever it returns.”
- “Kick off the invoice-intake scenario for the attached PDF, then tell me whether the router sent it down the Xero branch or the QuickBooks branch.”
- “Which of my on-demand scenarios could push this list into Airtable? Pick one and run it.”
Claude selects the right scenario from your library, executes it, waits on the Return output module, and reasons over the result. For flows that need human judgment in the middle — review this, then run that — it’s genuinely useful.
”On demand” is doing a lot of work in that phrase
The scheduling mode Make requires is also the integration’s honest label: these scenarios run when a person demands them, and the person is you, in a chat. A webhook that fires at 2 a.m. still executes your Make scenario the way it always has — but Claude takes no part in that run. There is no configuration where a Make event opens a Claude conversation, applies Claude’s judgment, and acts on the outcome. If you want Claude’s reasoning in the loop, you are the loop.
Two more Make-specific realities. Every run Claude kicks off consumes operations from your Make quota exactly like a normal execution, so a chatty session that calls three scenarios meters like three scenario runs. And while management scopes technically permit an MCP client to modify scenario contents, nobody should assemble a fourteen-module scenario with routers and error handlers through a chat window — the canvas remains where scenarios actually get built.
One tool instead of two: Carly
Running Claude-plus-Make means maintaining both halves of a relay: modules and operations on one side, prompts on the other, with you carrying results between them. For assistant-style work — email triage, follow-ups, meeting prep, recurring reports — Carly collapses that stack into one system that has its own triggers.
Carly watches for events itself: an email arriving, a calendar change, a Friday-morning schedule. When something fires, the whole flow runs in the cloud — your laptop can be closed. You never map a module: describe the job in plain English (“every Friday, pull unpaid invoices and email me a summary”), and Carly interviews you about the details, then assembles the workflow with you. It also finishes the job, sending email through Gmail or Outlook with attachments, updating tasks, and filing meeting notes rather than handing you output to relay.
AI agents start at $35/month, and steps that don’t use AI run free and unlimited — no per-operation meter ticking on every module. Carly connects to 200+ tools across 40+ categories; see integrations and the best AI workflow automation tools.
Claude’s Make connector vs Carly
| Claude (Make connector) | Carly | |
|---|---|---|
| Run a scenario when you ask in chat | Yes (On Demand scenarios) | Yes |
| React to a webhook, schedule, or new email on its own | No | Yes |
| Build the workflow from a plain-English description | No — canvas work in Make | Yes — interviews you, then builds it |
| Setup | Built-in connector + per-scenario config in Make | One-click |
| Cost model | Make operations per run + Claude plan | From $35/mo; non-AI steps free and unlimited |
| Runs while you’re offline | No | Yes (cloud) |
Frequently Asked Questions
Does Claude integrate with Make?
Yes — natively, as of 2026. Make is a built-in connector in Claude’s directory: enable it, complete the OAuth flow, and Claude can run any scenario you’ve set to On Demand scheduling with a Return output module. Guides claiming Make needs a custom MCP server describe the pre-2026 situation.
Can a Make webhook or schedule trigger Claude?
No. Claude’s connectors only operate inside a conversation you start. Make events run their scenarios as usual, but nothing on Make’s side can open a Claude chat or invoke Claude’s reasoning. Trigger-driven AI is what agent platforms like Carly exist for.
Which scenarios can Claude see?
Only active scenarios with On Demand scheduling in the organization you authorized, and they should end with a Return output module so Claude receives a result. Regular scheduled or webhook-triggered scenarios don’t appear as tools.
Does using the connector cost extra?
The connector itself is free to enable, but every run Claude triggers consumes Make operations from your plan’s quota like any other execution — and you’re on a Claude plan for the chat itself.
What if the workflow should run without me in the chat?
Then the connector is the wrong shape: it executes only when you prompt it. Carly fires on events and schedules 24/7 in the cloud, builds the flow with you in plain English, and sends the results itself. AI agents start at $35/month.
More: Claude connectors · Claude vs Carly · Can Claude send emails · Claude + Mailchimp · Best AI workflow automation tools · Claude + Bubble · Claude + OpenAI · Claude + Hugging Face
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."
