Drop the Dialerr softphone into any website as a floating widget. Your agents make and receive real calls without leaving the page โ no telephony stack to build.
Same widget, your choice of workflow. All point at the same secure, per-domain key.
Paste before </body>. Nothing to install or run โ it's the official Dialerr snippet, loaded from dialerr.com and sandboxed in an iframe. Your pk_ is publishable (safe in client code, like Stripe's). Drop data-token-url to let users log in inside the widget.
No key needed to start โ it opens a browser, you approve once, and it provisions a key and wires everything up (publishable key + secret land in your env; nothing sensitive on the command line). Already have a key? Add --key pk_.... Detects Next.js / Vite / CRA / HTML. Built for AI coding assistants โ see llms.txt.
Then just tell your assistant: โAdd a phone dialer to my app with Dialerr.โ It provisions a key for you (you approve once in your browser โ sign in or create an account there) and installs the widget. Works with Claude Code, Claude Desktop, Cursor, and any MCP client. The server only talks to dialerr.com โ it reads no files and runs no commands.
Typed, no build step, React as an optional peer.
SDKs give you power and weeks of build. Seat-based widgets work day one but chain every agent to a vendor account. Dialerr doesn't make the trade.
Twilio/Telnyx/Vonage hand you a Device object โ you build the UI, token server, and lifecycle yourself. Dialerr is a finished widget: paste one line, dial.
RingCentral/Zoom/Aircall embeds require every agent to hold that vendor's paid seat and log into it inside the widget. Dialerr signs agents in with a token minted by your backend โ no second login, no per-seat coupling.
Telephony MCP servers expose API actions and assume you have keys. Dialerr's MCP is the only one that provisions the key itself โ approve once in a browser, and your assistant installs the rest. See the full comparison โ
A single script tag, CLI command, or React component. No SDK wiring.
Browser-native inbound & outbound calling, SMS, voicemail โ no plugins.
Next.js, React, Vite, Webflow, WordPress, or plain HTML. It runs in a sandboxed iframe.
React to call_connected, call_ended, sms_received and more in your own UI.
Publishable + secret keys, locked to your domains. Short-lived tokens minted on your server.
Ships from your account. Update once, every embed gets it โ nothing to redeploy on your side.
In Dialerr โ Settings โ API Access, add a key and list the domains allowed to embed it.
Paste the script, run the CLI, or drop the React component. Add a tiny token route for seamless sign-in.
The ๐ launcher appears. Agents dial from your page; call events flow back to your code.
Add one route on your backend that mints a short-lived token with your secret key. The secret never touches the browser.
This page embeds the actual Dialerr toolbar with the very snippet above. Look bottom-right โ click the ๐ launcher.
Get your key โpk_live_xxx for a real publishable key and allowlist your domain in Settings โ API Access. Your site must be served over HTTPS for the microphone.