What is Gemini 3.8 Flash-Lite TTS?
Gemini 3.8 Flash-Lite TTS is the cheaper, faster of Google's two current text-to-speech models, and the one Google names as the drop-in upgrade from Gemini 3.1 Flash TTS, now a legacy preview. It takes the same requests as Gemini 3.8 Flash TTS, so switching is a change of model name, and it has the same voice library, voice design and consented voice replication.
What it gives up is breadth and some acting range: it speaks 101 languages against more than 130, and Google steers audiobooks, heavy vocal effects and dialect work to Flash TTS. It is a stable model with no shutdown date.
What changed from Gemini 3.1 Flash TTS?
A lower audio price and new voice tools, but a stricter script format that older prompts need changing for.
- Price
- $12 per million audio tokens at Standard rates, against $20 for 3.1 Flash TTS, with text input unchanged at $1.
- Voices
- Voice design, consented voice replication and the 2,000-voice library are new; 3.1 Flash TTS had neither design nor replication.
- Scripts
- Text is now read word for word, so stage directions and speaker labels move into a separate speech_metadata field.
- Output and tiers
- Unary requests return WAV instead of raw PCM, and Flex, Priority and context caching are now available.
Who is Gemini 3.8 Flash-Lite TTS best for?
Gemini 3.8 Flash-Lite TTS is best for voice agents and phone bots, read-aloud features, bulk dubbing, and apps moving off Gemini 3.1 Flash TTS.
Voice agents and phone bots
It streams raw audio while it is generated, and can return mu-law or A-law at 8 kHz for telephone lines.
Read-aloud features
Everyday single-speaker speech costs about $0.003 per 10 seconds at Standard rates.
Bulk dubbing and audio content
Batch and Flex halve the price to $6 per million audio tokens, about $0.0015 per 10 seconds.
Apps still on Gemini 3.1 Flash TTS
Google names it as the drop-in upgrade, and its $12 audio price is below the $20 that 3.1 Flash TTS charges.
What's included with Gemini 3.8 Flash-Lite TTS?
The same request format and voice tools as Flash TTS, a library of more than 2,000 voices, style control with inline tags, streaming and phone formats, and Batch, Flex and Priority tiers.
- Same API as Flash TTS
- Both 3.8 speech models take one request format, so you can move a feature up to Flash TTS by changing the model name.
- Voice design and replication
- Create a voice from a written description, or copy one from a 10 to 30 second clip with the speaker's recorded consent.
- Voice library
- 30 studio voices plus more than 2,000 others, filterable by language, accent, gender, pitch and use.
- Style field and inline tags
- Delivery goes in a per-line style field, and tags such as <laugh> or <short pause> place sounds at an exact word.
- Streaming and phone formats
- Streams raw 24 kHz audio, and returns WAV, mu-law or A-law at rates down to 8 kHz.
- Batch, Flex and Priority
- Batch and Flex halve the price, Priority costs 1.8 times Standard, and context caching is supported.
- SynthID watermark
- Google marks every clip its Gemini audio models generate with an inaudible SynthID watermark.
How much does Gemini 3.8 Flash-Lite TTS cost?
Gemini 3.8 Flash-Lite TTS is free on the Gemini API's rate-limited free tier; paid use costs $1 per million text tokens in and $12 per million audio tokens out, about $0.003 per 10 seconds of speech, or half that on Batch or Flex.
- Free, rate-limited text input and audio output
- Your content may be used to improve Google's products
- No Batch or Flex jobs
- Text input, audio output
- Half the Standard rate, about $0.0015 per 10 seconds of speech
- Cached input $0.125 on Batch, $0.05 on Flex
- Text input, audio output; about $0.003 per 10 seconds of speech
- Cached input $0.25, plus $1.00 per million tokens per hour of storage
- Paid-tier content is not used to improve Google's products
- Text input, audio output
- About $0.0054 per 10 seconds of speech
- Cached input $0.45
Prices are for the Gemini Developer API in US dollars. Audio counts as 25 tokens per second, so Standard works out to about $1.08 per hour of speech. Paid use needs a linked billing account and at least $5 of prepaid credit, which expires after 12 months. Google AI Studio is free to use, and prices on Gemini Enterprise Agent Platform can differ. Full pricing on ai.google.dev. We checked these prices on .
How do readers rate Gemini 3.8 Flash-Lite TTS?
No one has rated Gemini 3.8 Flash-Lite TTS yet. Be the first.
Sign in to rate. One rating per account.
Gemini 3.8 Flash-Lite TTS pros and cons
It gives you Flash TTS's voices and controls at two thirds of the audio price, but it speaks fewer languages and early users still report uneven voice consistency.
Pros
- Hacker News users singled out how little the 3.8 speech models cost, and Flash-Lite is the cheaper one: about $1.08 an hour of speech at Standard rates, or about $0.54 on Batch.
- It keeps Flash TTS's voice tools, including voice design, consented cloning and the 2,000-voice library, at two thirds of the audio price.
- Google reports it second only to Flash TTS on Hume AI's Overall Quality Index.
Cons
- It speaks 101 languages, 29 fewer than Flash TTS; Finnish, Swedish, Thai, Lithuanian, Slovenian and Swahili are among the gaps.
- Early Reddit users of the 3.8 speech models say voice consistency is still uneven, with a long text split into parts coming back in slightly different voices.
- One request holds at most two speakers, both prebuilt voices, so dialogue with designed or cloned voices has to be generated turn by turn.
- Voice replication in Google AI Studio is not offered in the EEA, UK, Switzerland, India, Illinois or Texas.
What are the best Gemini 3.8 Flash-Lite TTS alternatives?
Three tools on Recaply do Gemini 3.8 Flash-Lite TTS's job. Chatterbox Turbo is open source, so you can run it on your own server.
Gemini 3.8 Flash TTSFreeChoose it when you need audiobooks, character acting, two-voice scenes or one of the 29 languages Flash-Lite lacks; it costs $18 instead of $12 per million audio tokens.
Sonic 3.6FreeChoose it when phone-bot latency matters most; Cartesia quotes under 90 ms and clones a voice from 10 seconds of audio, with a $5-a-month plan covering about 133 minutes, but it speaks 44 languages.
Chatterbox TurboOpen SourceChoose it when you want to host English speech yourself for free; the 350M-parameter model is MIT-licensed and clones a voice from 5 seconds of audio.
Questions about Gemini 3.8 Flash-Lite TTS
What is Gemini 3.8 Flash-Lite TTS?
Gemini 3.8 Flash-Lite TTS is Google's lower-cost text-to-speech model for voice agents, read-aloud features and bulk dubbing, with the same voices and controls as Flash TTS in 101 languages.
Can Gemini 3.8 Flash-Lite TTS be used for phone calls?
Yes. It can return 8-bit mu-law or A-law audio, the formats phone systems use, at sample rates such as 8,000 Hz, and streaming requests send audio in chunks while it is still being generated.
Is Gemini 3.8 Flash-Lite TTS free?
Yes. Gemini 3.8 Flash-Lite TTS has a free Free tier plan: free, rate-limited text input and audio output.
How much does an hour of speech cost with Gemini 3.8 Flash-Lite TTS?
About $1.08 at Standard rates, from Google's price of $0.003 per 10 seconds, and about $0.54 on Batch or Flex. Text input adds $1 per million tokens, which is small next to the audio.
Is there an open source alternative to Gemini 3.8 Flash-Lite TTS?
Yes. Chatterbox Turbo is an open source tool that does the same job, so you can run your own.
What are the best Gemini 3.8 Flash-Lite TTS alternatives?
The closest alternatives on Recaply are Gemini 3.8 Flash TTS, Sonic 3.6 and Chatterbox Turbo.
Where can I try Gemini 3.8 Flash-Lite TTS?
On its official website, ai.google.dev.
Gemini 3.8 Flash-Lite TTS vs Flash TTS: which should I use?
Start with Flash-Lite TTS for voice agents, read-aloud and bulk jobs, and move to Flash TTS for audiobooks, heavy acting, two-voice scenes, dialects or a language Flash-Lite lacks. Both take the same requests, and Flash TTS costs $18 per million audio tokens against $12.
Which languages does Gemini 3.8 Flash-Lite TTS not support?
It covers 101 languages. Among those only Gemini 3.8 Flash TTS supports are Finnish, Swedish, Thai, Lithuanian, Slovenian, Swahili, Somali and Burmese, so check Google's list for your languages before you build.
Does Google use my text to improve Gemini 3.8 Flash-Lite TTS?
On the free tier, Google says your content is used to improve its products. On the paid tier it is not, so use a paid key for customer calls or anything private.
Our verdict
Gemini 3.8 Flash-Lite TTS is the Gemini voice model most apps should start with.
It has the same voices, voice design and cloning as Flash TTS at two thirds of the audio price, and Google built it for streaming voice agents and bulk jobs.
- Skip it for
- Audiobooks, character acting, two-voice scenes and the 29 languages it lacks, where Flash TTS fits better.
- Check first
- Make sure your languages are on its list of 101 before you build.
Gemini 3.1 Flash TTS
Gemini 3.1 Flash Live
Gemini 3.8 Live
Voxtral 3
Eleven v4
MAI-Voice-2