Where to?
Paste a URL, get a short link. No account needed.
Where to?
Connect it to the Claude or ChatGPT you already use, then just ask how your links are doing — it answers from your real campaign data.
Get started freeHow it works
Add it as a connector in Claude or ChatGPT. One-time, no code.
Everything you’ve shortened — in the dashboard, any time — is now readable by your AI. And you can make new links right in the chat.
“How did the summer campaign do?” It answers from your real click data — channel, timing, what’s working.
Connect whichever assistant you’re already in. Claude is the quickest; ChatGPT needs one extra step — turning on developer mode.
Search for toui and pick toui URL Shortener, or open claude.ai/directory/toui directly.
The toui sign-in page opens. Prefer adding it by URL? A custom connector pointed at https://mcp.toui.io works the same way.
toui lists what Claude will be able to do before you allow it. Afterwards, under Settings → Connectors → toui, set shorten_url to Always allow so it won’t ask every time.
Settings → Apps → Advanced. ChatGPT shows an “elevated risk” notice — that’s normal for any connector you add yourself, not specific to toui.
Connection = Server URL https://mcp.toui.io, Authentication = OAuth. Tick “I understand and want to continue” to enable Create.
toui’s own screen lists what ChatGPT will be able to do: create short links, read stats, and edit your links and bio page. Click Allow.
Then just say “shorten this with toui” — name toui so it calls the connector, and a toui.io link comes back in the chat.
Connected, your links become something you can just talk to — ask in plain language and toui answers from your real click data, no dashboard to dig through.
The connector is the no-setup path. But everything toui does is also a REST API with a small, zero-dependency SDK — the same shortening and stats, callable straight from your own code, free on every plan.
Read the developer docstoui connects through an official OAuth connector, so you never paste an API key into your AI — where it would be a real security risk. We want both: convenient, and safe.