ChatGPT + Ramp: What the Official Plugin Can Do in 2026
Yes, ChatGPT integrates with Ramp. There’s an official Ramp plugin in ChatGPT, built on Ramp MCP, Ramp’s own Model Context Protocol server. Install it once under Settings → Plugins and you can work your company’s spend in plain English: pull unsubmitted expenses, check card limits, approve or reject a reimbursement, lock a card. Ramp’s tools span cards, expenses, reimbursements, travel, and approvals, so this is one of the more action-capable finance plugins, not just a reporting window. But like every plugin in ChatGPT, it works when you or a schedule prompt it: nothing reacts when a transaction posts or an approval stalls.
Here’s what the ChatGPT Ramp integration actually does, how to turn it on, and what to use when you want spend management that runs without you.
What ChatGPT can actually do with Ramp
- Pull live spend data. “What did we spend on software last month, by department?” — answered from your real Ramp transactions, with merchant and category detail, no CSV export.
- Work the approval queue. Approve or reject transactions and reimbursements from the chat — “approve all pending reimbursements under $100 from the sales team” — instead of clicking through each one.
- Manage cards conversationally. Lock, unlock, or activate cards, and check limits. Someone’s card gets skimmed on a trip? “Lock Claire’s card” handles it mid-conversation.
- Edit transactions and create trips. Fix memos and categories, or set up travel — the operational cleanup work that usually eats a finance team’s Friday.
- Run inside agent sessions. With ChatGPT Work (launched July 9, 2026), you can @-mention Ramp and let an agent work across it and the rest of your connected stack for a long, metered run — a full month-end expense cleanup pass, say. Still a run you start or schedule.
How to set it up
- Check access on both sides. ChatGPT’s plugin directory is open across plans, though availability depends on your plan and workspace settings. In Ramp, most company-wide tools require Admin or Business Owner permissions.
- In ChatGPT, open Settings → Plugins, select Ramp, and click Install plugin.
- Authorize your Ramp account when prompted. The plugin runs on Ramp’s hosted MCP server, so there’s nothing to deploy or configure. (Developers can also connect via the Ramp CLI for Claude Code, Codex, and Cursor.)
- Ask a spend question (“show me this week’s pending approvals”) or invoke it explicitly with @Ramp in a prompt.
The limits that actually matter
- It doesn’t run on triggers. There’s no “when a transaction over $2,500 posts, flag it” or “when a reimbursement sits unapproved for three days, nudge the manager.” ChatGPT Work’s event triggers cover new Gmail messages, Slack channel messages and GitHub pull requests, not Ramp events, so ChatGPT touches Ramp only when you or a schedule prompt it.
- Bill approvals stay in Ramp. Ramp’s docs say bill approvals aren’t available through MCP yet, so those still require the Ramp interface. It doesn’t pay vendors either. The one spending path is an Agent Card: a credential scoped to one merchant and amount, issued right before a single approved checkout.
- Permissions gate most of it. The useful tools require Admin or Business Owner access, so this is a finance-team power tool more than something every employee connects.
- Scheduled at best, and metered. A ChatGPT Work task can call Ramp on a schedule, but runs draw on your plan’s allowance and nothing reacts when a transaction posts. It’s an errand, not a standing watch on company spend.
If you want Ramp work that runs on its own: Carly
The moment you want something to happen off a spend event — a Monday-morning spend summary by department in your inbox, a Slack ping to the founder when a transaction over $2,500 posts, a weekly list of unsubmitted expenses nagged to the right people — you’ve crossed past what a plugin in a chat is for.
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 code:
- Fires on events and schedules, 24/7, in the cloud. Large transaction posts, Monday 8am, end of month — Carly acts without a chat open.
- No-code setup. Tell Carly “every Monday, email me a spend summary by department” in plain English; it interviews you and builds the workflow.
- Connects spend to the rest of your work — Ramp data flowing into email, Slack, spreadsheets, and your accounting stack in one flow.
- Actually sends — drafts and sends email across Gmail and Outlook, posts to Slack, updates sheets and tasks.
- Connects to thousands of apps, Ramp included. An admin creates an app under Company → Developer in Ramp with the Client Credentials grant, then pastes the client ID and secret under Custom API Keys at carlyassistant.com/integrations, and Carly can read and act on your Ramp data in workflows.
AI agents start at $35/month, and steps in a workflow that don’t use AI run free. See integrations for the full list.
ChatGPT vs Carly
| ChatGPT (Ramp plugin) | Carly | |
|---|---|---|
| Live spend and transaction queries | Yes | Yes |
| Approve/reject from chat | Yes | Yes, in a workflow you define |
| Weekly spend digest, unprompted | Via a scheduled Work task | Yes, on a schedule |
| Flags a large transaction by itself | No | Yes, on any trigger |
| Runs without a session open | Scheduled Work tasks only (metered) | Yes (cloud, 24/7) |
| Pushes results into inbox / Slack / sheets | No | Yes |
| Emails the summary to your team | No | Yes (Gmail + Outlook) |
| Setup | Install the plugin | Describe it in plain English |
| Pricing | Your ChatGPT plan (availability varies) | AI agents from $35/mo |
ChatGPT’s Ramp plugin is a finance operator you drive from a chat. Carly is an assistant that watches your spend and acts while you’re doing something else.
Frequently Asked Questions
Does ChatGPT work with Ramp?
Yes. Ramp has an official plugin in ChatGPT, built on Ramp’s hosted MCP server, with tools across cards, expenses, reimbursements, travel, and approvals. Install it under Settings → Plugins. Most company-wide tools need Admin or Business Owner permissions in Ramp.
Can ChatGPT pay bills or move money in Ramp?
Not bills. Bill approvals aren’t available through Ramp MCP yet and still require the Ramp interface, and it doesn’t pay vendors. The one spending path is an Agent Card, a credential scoped to one merchant and amount for a single approved checkout. Otherwise it approves or rejects transactions and reimbursements, locks, unlocks and activates cards, edits transaction details, and creates trips, all within Ramp’s own permission controls.
Can ChatGPT alert me when a big transaction posts in Ramp?
No. ChatGPT reaches Ramp when you ask or on a schedule you set. Its event triggers don’t include Ramp, so it doesn’t watch for transactions, approval bottlenecks, or policy violations. For “when X happens in Ramp, do Y across my stack,” you need a trigger-based assistant like Carly, which connects to Ramp with your own API key.
How do I connect ChatGPT to Ramp?
Open Settings → Plugins in ChatGPT, select Ramp, click Install plugin, and authorize your Ramp account. The plugin runs on Ramp’s hosted MCP server, so there’s nothing to host. Then ask about your spend or invoke it with @Ramp.
More: ChatGPT email assistant · ChatGPT personal assistant · ChatGPT MCP · Can ChatGPT send emails · Claude + Ramp · 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."
