Gemini 3.8 Flash TTS supports 130+ languages: multilingual ad reads
Google lists 130+ languages for Gemini 3.8 Flash TTS and 100+ for Flash-Lite, with auto detection. How to use that for ad voiceover, and the Sume gap.

Gemini 3.8 Flash TTS (gemini-3.8-flash-tts) is documented with more than 130 languages, the largest count among the voice models I read this week; Flash-Lite TTS lists more than 100. Both detect the language automatically. For a multilingual ad campaign that means one API and one set of voices can cover many markets. It does not mean every language sounds equally good, and a language count is not a quality score.
What Google states
| Item | Gemini 3.8 Flash TTS | Gemini 3.8 Flash-Lite TTS |
|---|---|---|
| Model id | gemini-3.8-flash-tts | gemini-3.8-flash-lite-tts |
| Languages | More than 130 | More than 100 |
| Language detection | Automatic | Automatic |
| Voices | 30 prebuilt voices, an extended voice library, Voice Design, Voice Replication | Same speech-generation page |
| Price | $0.50 per 1M input tokens (through Dec 31, 2026), $9.00 per 1M output tokens | $0.50 in (promo), $6.00 per 1M output tokens |
A workflow for a multi-market ad
- Localise the copy first. Translating an ad word for word usually breaks rhythm and the call to action. Have a native writer adapt it.
- Pick one voice per market from the prebuilt set and keep it fixed across the campaign.
- Write numbers, prices, brand names and URLs the way you want them read. Auto detection can guess wrong on a short line that mixes scripts, so test those lines.
- Render each language, have a native listener check it, and keep a pass/fail log per line.
- Check the output length against the slot in the edit before you commit the schedule.
Where Sume stands
Sume does not ship Gemini TTS. Its TTS Router serves Cartesia Sonic only, so any market Sonic does not cover cannot be served from the TTS catalog; check the language list in the catalog (GET /v1/tts-router/models) before you plan a market. A Sume voice has a primary language and the API returns 409 tts_voice_language_mismatch when a request's language disagrees, so a copy-paste of the wrong language does not reach a paid render without a confirmation.
So for a 100-plus-language campaign, use Google directly. For a handful of major markets that Sonic covers, with the render, billing and receipts in one place, Sume is the managed alternative.
Sources
Related posts
More in Use cases
- Translated text in an image runs longer: check fit before you publish
When you translate the text in an image into a longer language, lines may not fit. A fit checklist for Ideogram 4.5 edits on Sume, and when to redraw in code.
- Gift guide by budget: three short Timeline renders beat one long one
Split a holiday gift guide into under-$25, $50 and $100 brackets, one short Timeline render each, and plan all three free first. Each is $0.10 a started minute.
- GitHub README banner with light and dark versions from Sume
Generate two 3:1 banners with a 1536x512 custom size, commit them to the repo, and switch them with a picture element and prefers-color-scheme in the README.
- Godot: import a Sume image as a texture, WebP or PNG
Godot imports PNG, JPEG and WebP. Drop a Sume image into res://, pick Lossless or VRAM Compressed, and use 1024x1024 where power-of-two sizes help.
Written by Sume