Your AI,
out loud.
Auris speaks every ChatGPT, Claude, Gemini, and Perplexity response the moment it starts streaming. Cook, walk, drive, rest your eyes. Stay in the conversation.
Reading was the bottleneck.
Speaks while it streams
No waiting for the full response. Auris starts talking the moment the first sentence lands, and keeps pace with the model.
Skips the noise
Code blocks, tables, and equations are skipped and announced. You hear ideas, not markdown syntax.
One habit, four AIs
ChatGPT, Claude, Gemini, and Perplexity. The same player, the same shortcuts, wherever you think.
Keyboard-first control
Alt+A reads the latest response or any selection. Alt+P pauses. Alt+S skips a sentence. Never touch the mouse.
On-device by design
No servers, no accounts, no analytics, no audio uploads. Auris reads your screen locally and that is the whole story.
Feels invisible
A quiet dot when idle, a glass player when speaking. Built to disappear into your workflow, not decorate it.
Get Auris today.
Browser extension
Reads ChatGPT, Claude, Gemini, and Perplexity aloud as they stream. Chrome Web Store listing is in review; founding users install directly in one minute:
1. Download and unzip.
2. Open chrome://extensions in your browser.
3. Turn on Developer mode (top right).
4. Click Load unpacked and pick the unzipped folder.
Desktop app
Lives in your menu bar. Reads AI desktop apps like Claude Desktop, and speaks any selected text in any app with Option+A.
1. Download and unzip.
2. Right-click Auris.app and choose Open the first time (it is a new app, so macOS asks once).
3. Grant Accessibility permission when prompted. Auris talks you through the rest.
Free while in beta. Both are 100% on-device: no account, no servers, nothing leaves your machine.
Founding access
- All four AI platforms
- Streaming voice, unlimited use
- Premium natural voices as they ship
- Founding badge + direct line to the maker
Questions
Is my data safe?
Yes, structurally. Auris has no backend. It runs inside your browser, reads only the conversation already on your screen, and uses your device's speech engine. Nothing is transmitted, stored, or logged anywhere. The extension requests the minimum possible permissions.
Does it break the AI platforms' terms?
Auris is an assistive layer, like a screen reader. It does not automate messages, scrape data, bypass limits, or touch any API. It simply voices content you already have on screen, on your own device.
Why not just use voice mode?
Voice mode is a separate conversation with a different feel and different limits. Auris voices your real chats, with your history and your long-form answers, in the interface you already use.
Which browsers?
Chrome and any Chromium browser (Edge, Brave, Arc) today. Firefox is on the roadmap.
What about code-heavy answers?
Code blocks are skipped and politely announced, so listening stays useful even for technical work. You can turn announcements off.