Read the web aloud
Select any paragraph or @-reference the current tab, and Tabbit reads it back to you with expressive, human-like voice.

Gemini 3.1 Flash TTS · Free in Tabbit · macOS & Windows
The AI-native browser that turns any web page, document, or chat into natural speech. No API key, no separate tab — just open Tabbit, pick Gemini 3.1 Flash TTS, and listen.
Features
Select any paragraph or @-reference the current tab, and Tabbit reads it back to you with expressive, human-like voice.
Gemini 3.1 Flash TTS covers global languages and regional accents, so your content sounds local wherever your audience is.
Use inline tags to control tone, speed, pauses, and even multi-speaker dialogue — no audio engineering skills required.
Let Agent Mode research, summarize, and then read the results aloud while you keep working in other tabs.
How it works
Launch Tabbit and choose Gemini 3.1 Flash TTS from the model picker.
Type text or press @ to reference the current page, a PDF, or a chat thread.
Hit play to hear it instantly, or save the audio for podcasts, voiceovers, and offline listening.
Comparison
Other paths force you to manage API keys or paste text into a playground. Tabbit puts the model inside your browser, ready to speak from any page.
| Gemini API | Replicate Playground | Tabbit Browser | |
|---|---|---|---|
| Price | Pay-per-use API billing | Pay-per-run cloud inference | Free during public beta |
| Google account / API key | Requires API key | Requires Replicate account | No API key needed |
| Read web pages aloud | Build your own integration | Paste text manually | @-reference tabs, PDFs & selections |
| Agent / automation | Code required | No automation | Agent Mode runs voice workflows in isolated tab groups |
| Model switching | Gemini TTS only | Limited model list | Gemini, GPT, Claude, DeepSeek, Qwen, and more |
| Desktop experience | Developer tools only | Web playground | Native macOS & Windows app |
Use cases
Turn long articles and newsletters into a personal podcast for your commute.
Give your eyes a rest. Listen to pages instead of reading them.
Generate voiceover drafts, ad reads, and narration in seconds.
Hear native accents, slow down speech, and practice pronunciation.
Let Agent Mode compile a briefing and read it aloud while you review sources.
Listen to one tab while you work, browse, or take notes in another.
FAQ
Yes. During the Tabbit international public beta, Gemini 3.1 Flash TTS is available at no extra cost alongside GPT, Claude, and other top models. No separate Google One or API subscription is required.
No. Tabbit routes model access through its own international infrastructure, so you can use Gemini 3.1 Flash TTS with only your Tabbit account.
Yes. Use @ to reference the current tab, a selected paragraph, or a PDF, then ask Tabbit to read it aloud. The browser context becomes the voice input.
Gemini 3.1 Flash TTS supports 70+ languages and a wide range of accents, styles, and speaker profiles. You can also use inline tags to adjust emotion, speed, and pauses.
Inline tags are markers inside the text that tell the model how to speak — for example <happy>, <whisper>, <pause>, or <speaker name="host">. Tabbit passes these tags directly to Gemini 3.1 Flash TTS.
Yes. After Tabbit generates the audio, you can play it inline or download it as an audio file for podcasts, voiceovers, or offline listening.
Yes. You can instruct Agent Mode to summarize a batch of pages and then read the summary aloud, turning research into a hands-free briefing.
Yes. Tabbit ships the same AI-native feature set — including Gemini 3.1 Flash TTS — on both macOS and Windows.

Download Tabbit for macOS or Windows and use Gemini 3.1 Flash TTS alongside GPT, Claude, and more — no invite code.