Gemini 3.8 TTS logo

gemini-3.8-flash-tts · gemini-3.8-flash-lite-tts

Gemini 3.8 TTS, in your browser

Expressive text-to-speech: 30 voices, single or two-speaker dialogue, accents, styles, pace and inline tone tags. MP3 and WAV downloads.

Model
Mode

google/gemini-3-8-flash-tts

Speaker 1voice
Voice · 30 options
Text0 / 5000
12 guest credits free - sign up for 60· Est. cost: 1 cr
Output

Your audio will appear here.

Local history · last 20

Generations you run on this device appear here. Signed-in history also lives in Library.

Released September 23, 2026

What's new in Gemini 3.8 TTS

Expressive by default

Flash targets creative work - audiobooks, dialogue and dialects - in 130 languages, with style, accent and pace controls per speaker.

Inline tone tags

Write [laughs], [whispers] or [short pause] directly in the text and the model performs them.

MP3 or WAV

Gemini 3.8 returns studio-quality WAV by default; grab the smaller MP3 for YouTube, TikTok and podcasts with one click.

Model picker

Flash vs Flash-Lite vs 3.1 Preview

3.8 Flash3.8 Flash-Lite3.1 Flash Preview
PositioningExpressive creative workLow-latency, cost-efficientLegacy preview (replaced)
Languages130101Fewer
Credits here / 1k chars2.121.42—
Speech per 100 credits≈ 52 min≈ 78 min—
Best forAudiobooks, dialogue, dialectsHigh-volume and real-time-ish useDeprecated path

Billed per 100 characters, rounded up to a whole credit per generation. Speech time assumes about 15 characters per second. Developer pricing for the Google API is on the API guide.

Pricing facts

Simple credit pricing

12

free guest credits

Try the playground without an account.

60

free credits on signup

About 30 minutes of speech on Flash (28k characters), 47 on Flash-Lite.

$1.40

per hour of speech, on yearly Pro

Monthly from $9.90, yearly or one-time packs. See pricing

FAQ

Questions

What is Gemini 3.8 TTS?
Gemini 3.8 Flash TTS and Flash-Lite TTS are Google's text-to-speech models released on September 23, 2026. Flash targets expressive creative work - audiobooks, dialogue and dialects - in 130 languages. Flash-Lite is the low-latency, cost-efficient option in 101 languages, replacing gemini-3.1-flash-tts-preview.
Is this site affiliated with Google?
No. Gemini 3.8 TTS is an independent site, not affiliated with Google. We provide an easy console on top of the Gemini 3.8 TTS models via an AI generation provider.
How much does it cost?
You pay in credits: 2.12 credits per 1,000 characters on 3.8 Flash and 1.42 on 3.8 Flash-Lite, billed per 100 characters and rounded up to a whole credit per generation - a short 200-character line costs 1 credit. Guests get 12 free credits; registered accounts get 60, about 30 minutes of speech. Paid plans start at $1.40 per hour of speech.
What are inline tone tags?
Gemini 3.8 TTS understands inline tags such as [laughs], [sighs], [whispers], [excited] and [short pause] placed inside your text. The tag chips under each text field insert them at the cursor.
Can I generate a two-person dialogue?
Yes. Two-speaker dialogue mode gives you a turn list: each turn picks Speaker 1 or Speaker 2, and you can add, remove and reorder turns. Each speaker gets its own voice, accent, style and pace.
What audio format do I get?
Download every generation as MP3 or WAV. Gemini 3.8 returns studio-quality WAV; the MP3 is converted in your browser, so nothing extra is stored.
Does this site offer voice cloning or the 2,000+ voice library?
Not yet. Google announced voice design, voice cloning and a 2,000+ voice library for Gemini 3.8, but the API this site runs on currently exposes the 30 prebuilt voices plus a free-text audio profile per speaker. We will add cloning and the larger library as soon as they are available.

Independent site, not affiliated with Google.