Gemini 3.8 Flash TTS
Gemini 3.8 Flash TTS is Google's text-to-speech model for creative direction and character voices, announced on 23 September 2026 with a cheaper sibling, Flash Lite TTS, a generation after the 3.1 Flash TTS preview. Google's own API adds a library of more than 2,000 voices, voice design from a text prompt and replication of your own voice from about 30 seconds of audio; fal exposes the 30 prebuilt voices, plain-language style instructions, inline vocal events like <laugh> and two-speaker dialogue. Google's model page lists more than 130 languages. On fal it costs $0.045 per thousand characters; on the Gemini API, $9 per million audio tokens through 31 December 2026 and $18 from 1 January 2027.
Anvisha Pai, Co-founder & CEO, Voyager
Google DeepMind · released
›What we asked for
Hi, I'm Gemini 3.8 Flash TTS, Google's voice model. I take stage directions written in plain language, but today I'm reading the script as written. Here is the same script every voice on this site reads, so you can compare us.
google/gemini-3.8-flash-tts
voice: Kore
returned 16.2 seconds of wav
At a glance
- Price
- $0.045 per thousand characters on fal, a little under Gemini 3.1 Flash TTS; our four scripts cost about four and a half cents.
- Voices
- Thirty prebuilt voices on fal, Kore by default, which we used; Google's own API adds a library of more than 2,000 and voice replication.
- Direction
- A style_instructions field in plain language, per turn or for the whole script, plus inline events like <laugh>; we sent neither.
- Dialogue
- Exactly two speakers, each with its own voice, over an ordered list of turns.
- Languages
- More than 130 on Google's model page, auto-detected.
Other Gemini TTS models: Gemini 3.8 Flash Lite TTS, Gemini 3.1 Flash TTS. Compare all 3.
Anvisha's take
We have not scored the four readings by ear; press play and judge the voice yourself. What we measured: Gemini 3.8 Flash read the four scripts in 15 to 23 seconds of audio each and returned in seven to nine seconds, about twice as fast as Gemini 3.1 Flash TTS, which took 13 to 19, for four and a half cents across all four. Its narration paragraph came back in 22.6 seconds, the shortest reading of it on the site and three seconds shorter than 3.1's; we have not checked word by word that nothing was dropped. It ran with Kore, the voice we gave 3.1, and no style instructions, so its headline feature, direction, is the one these scripts withhold.
Gemini 3.8 Flash TTS real output: the same 4 scripts every model gets
Every voice model on this site gets the same 4 scripts, so you can compare like with like. These are the readings Gemini 3.8 Flash TTS returned, one try each, nothing re-rolled or retouched. Generated on . The 4 readings took 7 to 9 seconds each and cost $0.0439 in total.
Pacing, warmth and whether the pauses land where a voice actor would put them.
›What we asked for
Meet the Ember mug. It keeps your coffee at exactly the temperature you choose, for up to ninety minutes, and it tells your phone when it's ready. Warm from the first sip to the last. Ember. Coffee, on your terms.
google/gemini-3.8-flash-tts
voice: Kore
returned 15.1 seconds of wav
Steady narration over a long paragraph, clean list rhythm and a natural close.
›What we asked for
A pour-over works in four steps. First, rinse the paper filter with hot water so it doesn't taste of paper. Second, add the grounds and pour just enough water to wet them, then wait thirty seconds while they bloom. Third, pour the rest of the water in slow circles. Fourth, wait about three minutes for it to drip through. That's it: a cup that tastes like the coffee, not the machine.
google/gemini-3.8-flash-tts
voice: Kore
returned 22.6 seconds of wav
Range: excitement, then a hushed aside, from the words alone with no tags.
›What we asked for
We did it! We actually did it! I can't believe it's finally over. Okay. Okay, deep breath. Let's not tell anyone until Monday, alright? Just... let me have this one quiet night.
google/gemini-3.8-flash-tts
voice: Kore
returned 16.3 seconds of wav
Reading numbers, times, currency, an email address and two hard names correctly.
›What we asked for
Your order ships on the 24th of September 2026 and arrives by 9:15 a.m. The total is $1,249.99, including VAT. Questions? Ask for Siobhan Nguyen or Ravi Parikh on 0800 555 0199, or email help@moda.app.
google/gemini-3.8-flash-tts
voice: Kore
returned 22.6 seconds of wav
How the cost under each picture was worked out: fal's Gemini 3.8 Flash TTS rate of $0.045 per 1,000 characters (fal model page and pricing API, read 2026-10-22) for the 213 characters sent; fal does not return the charge.
Compare Gemini 3.8 Flash TTS
Gemini 3.8 Flash TTS pricing
| Where | Unit | Price | Source |
|---|---|---|---|
| fal | 1,000 characters Billed on the spoken text; style instructions and speaker names are free. | $0.045 | fal model page and pricing API, read |
| Gemini API | Million audio output tokens Through 31 December 2026, then $18. Text in is $0.50 per million, then $1. Audio is 25 tokens a second. | $9.00 | Gemini API pricing, read |
Is Gemini 3.8 Flash TTS free?
Google's pricing page lists Gemini 3.8 Flash TTS as free of charge on the Gemini API's free tier, which is rate-limited. fal charges per character.
- •Gemini API free tier: input and output free of charge.
- •Gemini API paid tier: $9 per million audio tokens through 31 December 2026.
- •fal: $0.045 per thousand characters.
Gemini API pricing, read
Where to use Gemini 3.8 Flash TTS
- Gemini API · Model id gemini-3.8-flash-tts; voice library, design and replication.
- Google AI Studio · Try it in the browser; voice replication is not offered in some regions.
- fal · $0.045 per thousand characters, 30 voices, WAV out.
- Voyager · Runs inside Voyager, the agent for creative work, with your own fal key.
Voyager
Run Gemini 3.8 Flash TTS inside Voyager
Voyager is an agent for creative work that runs every model in this reference, Gemini 3.8 Flash TTS included. Brief it, and it picks the model, makes the thing and hands back something you can edit. Create an account to get started.
What we noticed running Gemini 3.8 Flash TTS
- •Returned in seven to nine seconds per reading, against 13 to 19 for Gemini 3.1 Flash TTS with the same voice.
- •22.6 seconds for the paragraph, the shortest reading of it on the site; Gemini 3.1 took 25.8 and Kokoro 22.9. See the explainer paragraph reading.
- •22.6 seconds again, the same length as the much longer narration script to the hundredth; the two files differ. See the names, dates and prices reading.
- •No style_instructions and no vocal events were sent; what you hear is the default delivery.
Gemini 3.8 Flash TTS limits and API parameters
| Limit | Value |
|---|---|
| Input | 8,192 tokens for the whole request, script and directions together. Gemini API model page, read |
| Output format on fal | WAV only; ours came back as 24 kHz mono. fal API reference, read |
| Cloning on fal | None; the 30 prebuilt voices only. Replication is on Google's own API. fal API reference, read |
›Every setting, for developers
fal (google/gemini-3.8-flash-tts)
fal API reference, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| prompt | string | The verbatim script, with optional vocal events like <laugh> or <sigh> Required for one speaker; omit it for dialogue. | |
| voice | enum | 30 voices including Kore, Puck, Charon, Zephyr, Aoede | Kore |
| style_instructions | string | Plain-language delivery, for example 'cheerful and friendly' Not billed. | |
| speakers | list | Exactly two speaker aliases, each with a voice | |
| turns | list | Ordered dialogue turns: speaker, text and an optional per-turn style | |
| enable_safety_checker | boolean | true, false Only authorised callers may turn it off. | true |
What Gemini 3.8 Flash TTS will not say
Google's Generative AI Prohibited Use Policy applies: no impersonation of a real person meant to mislead, no fraud or scams, no harassment, no sexual content involving minors. Google says every clip from its Gemini audio models carries a SynthID watermark. Voice replication, on Google's own API and not through fal, needs a consent recording from the voice's owner.
Alternatives to Gemini 3.8 Flash TTS
Frequently asked questions
What is Gemini 3.8 Flash TTS?
Google's text-to-speech model for directed, expressive speech, announced on 23 September 2026, with style instructions, two-speaker dialogue and more than 130 languages. It is on the Gemini API, Google AI Studio and fal.
How much does Gemini 3.8 Flash TTS cost?
$0.045 per thousand characters on fal, or on the Gemini API $9 per million audio tokens and $0.50 per million text tokens through 31 December 2026, doubling on 1 January 2027. Read on 22 October 2026.
Can Gemini 3.8 Flash TTS clone a voice?
On Google's own API, yes, from about 30 seconds of audio with a consent recording from the voice's owner. Through fal, no: it offers the 30 prebuilt voices.
How this page is made
Prices, limits and settings are read from the linked vendor and host pages on the dates shown. Every picture was generated by us through fal with the request shown under it, and the original files are kept unedited. A generation that failed is shown as a failure. Model and vendor names are used to identify the products; no vendor imagery appears on this page.
