NEWGemini 3.8 Flash TTS is live in the studio

Text to speech you can direct.

An AI text to speech studio where you decide how it sounds. Type your text, pick a voice, and add a laugh or a pause where you want one. Sign up free for 10,000 characters a month.

languages
130+
studio voices
30
voices per podcast
2
Type or paste your text209 / 5,000
Add a moment:It's performed, not read out.
Pick a voice
Tone (optional)
0:00 / 0:00

What people make

What You Can Make with AI Text to Speech

Each job uses a different part of the AI voice model. Pick one and the studio sets up the voices and tone for it.

Voiceover for video

Explainers, shorts and course lessons. Pick a voice, set the tone once, and fix single lines without re-recording.

YouTubers · course creators

tone presets + inline tags

Two-host podcast

Use it as an AI podcast generator: write a conversation, give each speaker a voice, and short reactions like |oh| land between the lines.

Podcasters · newsletter writers

2 voices per request

Characters from a sentence

An AI voice generator that starts from a sentence: describe a voice in plain words and hear it right away. Reuse the description for every line of your game or story.

Game devs · storytellers

voice design from a description

Stories and audiobooks

Paste a chapter or write a full-cast scene. Speakers stay distinct, and pauses and sighs land where you put them.

Authors · storytellers

up to 8,000 characters per request

Learning and dubbing

Pronunciation for flashcards, or a lesson read in another language. The language is detected from the text.

Students · educators

130+ languages, auto-detected

Short-video stories

Punchy narration for vertical video, with a laugh or a beat exactly where the joke lands.

Short-form creators

energetic tone preset

Pricing

AI Text to Speech Pricing: Pay per Character, No Subscription

No subscription and no auto-renewal. One minute of English speech is about 900 characters.

Free

$0

10,000 characters a month

Try everything with no card.

  • Sign in with Google, no card needed
  • Up to 5,000 characters per request
  • All 30 voices
  • Voiceover, podcast and new-voice modes
Start free
Popular

Creator pack

$9

300,000 characters

For regular voiceovers and podcasts.

  • Up to 8,000 characters per request
  • Everything in Free
  • Buy again any time

Studio pack

$29

1,000,000 characters

For long projects and teams that publish weekly.

  • Up to 8,000 characters per request
  • Everything in Free
  • Best price per character

Questions

Text to Speech FAQ: Voices, Languages, Limits & Licensing

Is ttsloom made by Google?
No. ttsloom is an independent studio that runs speech models, including Gemini 3.8 Flash TTS. It is not affiliated with or endorsed by Google.
What do tags like <laugh> do?
They are performed, not read out. A laugh, a sigh or a pause changes the delivery at that spot. Words between bars, like |oh hmm|, are spoken as a quick reaction.
How long does it take?
Usually around 20 seconds, sometimes longer when the service is busy. If something fails, you are not charged.
Which languages work?
The flash model supports 130+ languages and the lite model 100+. The language is detected from your text.
How many characters can I generate?
After you sign up you get 10,000 free characters every month. One request can be up to 5,000 characters on the free plan and up to 8,000 with a character pack.
Can I use the audio in my projects?
Yes. See the Terms for the details of what you can do with generated audio. Read the Terms
Is my text stored?
To make the audio, your text is sent to the speech model provider. The generated audio is deleted automatically after 30 days. Read the Privacy Policy
Can I clone my own voice?
Not yet. Voice cloning needs a consent check, so we're treating it separately. Tell us if you need it.
Is there an API?
Not yet. If you'd build something with one, send us a note through the contact page.