ChatGPT + tl;dv: What the Integration Can (and Can't) Do in 2026
Yes, with some setup — there’s no tl;dv app in the ChatGPT apps directory, but tl;dv ships an official MCP server that exposes your whole meeting library — recordings, transcripts, and AI highlights across Zoom, Google Meet, and Microsoft Teams — to any MCP client. The code lives on GitLab, you run it yourself via Docker or Node, and it authenticates with a tl;dv API key from Personal Settings → API Keys — which means a tl;dv Business or Enterprise plan, since that’s where API access lives. One wrinkle for ChatGPT specifically: custom connectors want a remote URL, so you’ll deploy the server somewhere reachable rather than running it on your laptop the way Claude Desktop allows.
Here’s what the ChatGPT tl;dv integration actually does, how to set it up, and what to use when you want the follow-up on those meetings to run without you.
What ChatGPT can actually do with tl;dv
- Query your meeting library in plain English. “What did Miguel say about the renewal timeline in last week’s calls?” — the server lists meetings with filters and pulls complete metadata.
- Pull full transcripts, platform-agnostic. tl;dv normalizes Zoom, Google Meet, and Teams transcripts into one consistent format, so ChatGPT reads them all the same way.
- Work from AI highlights. The server exposes tl;dv’s AI-generated key moments, so you can ask for the decisions and objections without feeding ChatGPT an hour of raw transcript.
- Draft follow-ups from real calls. Paste-free follow-up emails, to-do lists, and summaries grounded in what was actually said — in the session you’re driving.
- Run inside agent sessions. With ChatGPT Work (launched July 9, 2026), an agent can @-mention your connectors and sweep a quarter’s worth of customer calls for churn signals in one long, metered run. Still a run you start.
How to set it up
- Be on a tl;dv Business or Enterprise plan, then generate an API key in tl;dv under Personal Settings → API Keys.
- Grab the official server from gitlab.com/tldv/tldv-mcp-server and run it with Docker or Node, with your API key in the environment.
- For ChatGPT, deploy it to a URL ChatGPT can reach (Fly.io or Render make this a ten-minute job) — custom connectors take a remote MCP endpoint, not a local process.
- In ChatGPT’s settings, add the deployed URL as a custom connector.
- Ask about your meetings (“summarize yesterday’s demo with the Segment team”) or @-mention the connector in a prompt.
The limits that actually matter
- Nothing happens when a recording lands. The thing everyone actually wants — “when tl;dv finishes processing a call, send the follow-up and file the action items” — isn’t what a chat connector does. ChatGPT queries tl;dv when you ask; it never fires on a new recording.
- Self-hosting is on you. tl;dv doesn’t run a hosted MCP endpoint, so the ChatGPT route means deploying and maintaining a small service, plus keeping the API key secure.
- Plan-gated. API keys — and therefore the MCP server — require tl;dv’s Business or Enterprise tier. Free and Pro users are out of this particular game.
- Session-bound, even in agent mode. ChatGPT Work runs are long but manually started and metered against your plan’s allowance — a research errand, not a standing pipeline off your call library.
- The output stays in the chat. ChatGPT can draft the perfect follow-up from a tl;dv transcript; it won’t send it, log the call in your CRM, or create the tasks.
If you want tl;dv work that runs on its own: Carly
To be clear: tl;dv is doing its job perfectly here — recording, transcribing, and distilling every call across Zoom, Meet, and Teams. The gap is downstream: every transcript that lands creates twenty minutes of follow-up work that a chat session only helps with when you remember to open one.
That’s where Carly fits — not as a replacement for tl;dv, but as the assistant that acts on what tl;dv captures:
- Fires when the transcript lands, 24/7, in the cloud. When tl;dv finishes processing a sales call, Carly drafts the follow-up email in Gmail, logs the call notes against the deal in HubSpot, and creates a Linear ticket for each commitment made on the call.
- Scheduled digests. Every Friday at 4pm, Carly sweeps the week’s tl;dv highlights across the team’s customer calls and emails a themes-and-objections digest to sales leadership.
- No-code setup. Tell Carly “after every tl;dv recording, post the highlights to #sales in Slack and update the CRM” in plain English; it interviews you and builds the workflow.
- 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 — and Carly natively integrates with tl;dv.
ChatGPT vs Carly
| ChatGPT (tl;dv MCP connector) | Carly | |
|---|---|---|
| Query transcripts and highlights | Yes | Yes |
| Draft a follow-up from a call | Yes, in the chat | Yes — and sends it |
| Acts when a new recording lands | No | Yes, on the trigger |
| Weekly call-themes digest, unprompted | No | Yes, on a schedule |
| Runs without a session open | No (agent runs are started + metered) | Yes (cloud, 24/7) |
| Logs calls to CRM, creates tasks | No | Yes |
| Setup | Self-host + deploy the MCP server | Describe it in plain English |
| Pricing | Paid ChatGPT plan + tl;dv Business | AI agents from $35/mo |
ChatGPT with tl;dv’s MCP server is a meeting-library analyst you question in a chat. Carly is the assistant that runs the follow-up the moment tl;dv finishes transcribing — with tl;dv doing what it does best underneath.
Frequently Asked Questions
Does ChatGPT work with tl;dv?
Yes, via MCP. tl;dv ships an official MCP server (on GitLab) covering Zoom, Google Meet, and Teams recordings. There’s no tl;dv app in the ChatGPT directory, so you run the server yourself — Docker or Node with your tl;dv API key — and add its deployed URL to ChatGPT as a custom connector.
What plan do I need for the tl;dv MCP server?
A tl;dv Business or Enterprise account. The server authenticates with a tl;dv API key generated under Personal Settings → API Keys, and API access is a Business/Enterprise feature.
Can ChatGPT summarize my tl;dv meetings automatically after each call?
No. ChatGPT only reads tl;dv when you prompt it in a session — it can’t watch for new recordings. For “when tl;dv finishes a call, send the follow-up and file the action items,” use a trigger-based assistant like Carly, which natively integrates with tl;dv.
Is tl;dv’s MCP server hosted, or do I run it myself?
You run it yourself — tl;dv publishes the server on GitLab with Docker and Node options, but doesn’t operate a hosted endpoint. For ChatGPT custom connectors you’ll need to deploy it to a reachable URL; Claude Desktop and Cursor can run it locally.
More: ChatGPT + Fireflies · ChatGPT + Fathom · ChatGPT + Zoom · ChatGPT MCP · Claude + tl;dv · What is ChatGPT Work
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."
