ChatGPT + Make: What the Integration Can (and Can't) Do in 2026
Yes, ChatGPT connects to Make — through Make’s official MCP server, which you add to ChatGPT as a custom app. There’s no Make app in ChatGPT’s built-in directory; the supported route in Make’s help docs is developer mode plus your MCP toolbox URL. Once it’s wired up, the scenarios you expose show up as tools ChatGPT can call: ask in plain English and ChatGPT runs your lead-enrichment scenario, your invoice sync, your Slack digest. The catch is the direction of travel — ChatGPT can run Make, but only in a session you’re driving. Between chats, nothing is listening.
Here’s what the ChatGPT Make integration actually does, how to set it up, and what to use when you want automation that builds and runs itself.
What ChatGPT can actually do with Make
- Run your scenarios on demand. Any scenario you set to on-demand and expose through the MCP toolbox becomes a callable tool — “run my HubSpot lead-enrichment scenario for these five contacts” just works.
- Pass real inputs. Scenario inputs surface as typed parameters, so ChatGPT can fill them from the conversation — a domain to enrich, a row to append, a customer email to look up.
- Reason over the results. Scenario output comes back into the chat, so you can chain: run the scenario, inspect what it returned, decide what to run next.
- See your whole toolbox. Every scenario you expose appears in the app’s tool list, and you can restrict which ones ChatGPT is allowed to call from the connection’s settings.
- Run inside agent sessions. With ChatGPT Work (launched July 9, 2026), you can @-mention connected apps and let an agent chain Make scenarios with the rest of your connected stack in a long, metered run — a quarter-end data cleanup, say. Still a run you start.
How to set it up
- Have a paid ChatGPT plan (custom apps aren’t on the free tier) and a Make account with at least one scenario set to run on demand.
- In Make, open your MCP toolbox and copy the toolbox URL and key.
- In ChatGPT on the web, go to Settings → Apps → Advanced settings and enable Developer mode, then back in Apps click Create app.
- Paste the full URL in the format
<mcp toolbox url>/t/<toolbox key>/stateless(Make recommends the stateless transport for reliability), set Authentication to no auth, and create the connection. - Ask ChatGPT to list the available tools, then invoke a scenario by describing what you want.
The limits that actually matter
- The key is the security model. The toolbox key lives in the URL and the connection uses “no auth” — treat that URL like a password, because anyone holding it can run your scenarios.
- ChatGPT runs scenarios; it doesn’t build them. The hard part of Make — designing the module chain, mapping fields, handling errors — still happens in Make’s visual editor. ChatGPT only calls what you’ve already assembled.
- It doesn’t watch anything. Make’s own webhooks and schedules keep firing, but ChatGPT never joins in on a trigger. “When a Typeform response lands, have the AI triage it” is not something this connection can do — ChatGPT only acts when you prompt it.
- Two meters running. Every call burns Make operations and ChatGPT usage (agent runs are metered against your plan’s allowance). A chatty session over a heavy scenario adds up on both bills.
If you want automation that builds and runs itself: Carly
Notice the shape of the work here: you build automation in one tool, then use a second tool to invoke it by hand. The moment you want the whole loop — described in plain English, triggered by real events, running while you sleep — you’ve outgrown the pairing.
That’s where Carly fits. Carly is an AI executive assistant that acts on triggers across your whole stack, set up by conversation instead of a module canvas:
- Fires on events and schedules, 24/7, in the cloud. New form response, inbound email, Monday 8am — Carly acts without a chat open.
- No-code setup, genuinely. Tell Carly “when a demo request comes in, enrich the lead, add it to HubSpot, and Slack the sales channel” — it interviews you and builds the workflow. No modules to wire.
- Plays well with Make. Keep the scenarios you love; Carly natively integrates with Make and can trigger scenarios as one step in a larger flow.
- Actually sends — drafts and sends email across Gmail and Outlook, updates your CRM, manages tasks.
- Connects to anything — 200+ native integrations, plus any other tool via your own API key.
AI agents start at $35/month, and steps in a workflow that don’t use AI run free and unlimited. See integrations.
ChatGPT vs Carly
| ChatGPT (Make MCP) | Carly | |
|---|---|---|
| Run an existing scenario from chat | Yes | Yes (as a workflow step) |
| Build the automation in plain English | No (build in Make’s editor) | Yes, by conversation |
| Reacts to a webhook or event by itself | No | Yes, on any trigger |
| Runs without a session open | No (agent runs are started + metered) | Yes (cloud, 24/7) |
| Auth model | Toolbox key in a URL | OAuth / managed connections |
| Sends email, updates CRM, manages tasks | Via scenarios you pre-build | Yes, natively |
| Setup | Developer mode + MCP URL | Describe it in plain English |
| Pricing | Paid ChatGPT plan + Make operations | AI agents from $35/mo |
ChatGPT plus Make is a remote control for automations you’ve already built. Carly is an assistant that builds and runs the automation for you.
Frequently Asked Questions
Does ChatGPT work with Make?
Yes, via Make’s official MCP server. There’s no Make app in ChatGPT’s directory — you enable Developer mode in ChatGPT’s app settings, create a custom app, and paste your Make MCP toolbox URL. Scenarios you set to on-demand then appear as tools ChatGPT can run.
Can ChatGPT build Make scenarios for me?
No. The MCP connection exposes existing on-demand scenarios as callable tools; designing the scenario — modules, field mapping, error handling — still happens in Make’s visual editor. ChatGPT can help you plan a scenario in conversation, but it can’t assemble one in your account.
Can ChatGPT trigger a Make scenario automatically when something happens?
No. Make’s own webhooks and schedules run independently, but ChatGPT only calls scenarios inside a session you start. For “when X happens, have an AI decide and act,” you need a trigger-based assistant like Carly, which fires on events and schedules in the cloud.
How do I connect ChatGPT to Make?
In ChatGPT’s web settings, enable Developer mode under Apps → Advanced settings, click Create app, and paste your Make MCP toolbox URL in the format <url>/t/<key>/stateless with authentication set to “no auth.” Make’s help center walks through each step.
More: ChatGPT email assistant · ChatGPT personal assistant · ChatGPT MCP · Can ChatGPT send emails · Claude + Make · Best AI workflow automation tools
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."
