New Gemini TTS models review

Open Forum
1

With the recent release of Google’s new Gemini 3.8 TTS lineup, I wanted to share a quick comparison across models and languages.

I think it can be utilized to make a full audiobook when there’s no human voice actor’s one.

image
image1275×773 124 KB

If you’re using my Add-on, you can generate the full lesson audio with these TTSs in the editor page.
This is the guide explaining how to do it.

1. English Samples

  • gemini-3.1-flash-tts-preview
  • gemini-3.8-flash-lite-tts
  • gemini-3.8-flash-tts

2. Japanese Samples

  • gemini-3.1-flash-tts-preview
  • gemini-3.8-flash-lite-tts
  • gemini-3.8-flash-tts

3. Korean Samples

  • gemini-3.1-flash-tts-preview
  • gemini-3.8-flash-lite-tts
  • gemini-3.8-flash-tts
3 Likes
2

Supported languages

Language Language Language
Acehnese (Arab script) Greek Nepali (individual language)
Afrikaans Guarani Nigerian Fulfulde
Akan Gujarati North Azerbaijani
Amharic Haitian Creole Northern Sotho
Armenian Halh Mongolian Northern Uzbek
Assamese Hausa Norwegian Bokmål
Awadhi Hebrew Norwegian Nynorsk
Balinese Hindi Nyanja
Bangla Hungarian Occitan
Banjar (Arab script) Icelandic Odia (individual language)
Banjar (Latn script) Igbo Pangasinan
Bashkir Iloko Persian (Afghanistan)
Basque Indonesian Polish
Belarusian Iranian Persian Portuguese
Bemba Italian Punjabi
Bhojpuri Japanese Romanian
Bosnian Javanese Russian
Buginese Kabyle Santali
Bulgarian Kamba Serbian
Burmese Kannada Sindhi
Cantonese Kashmiri (Arab script) Sinhala
Catalan Kashmiri (Deva script) Slovak
Cebuano Kazakh Slovenian
Central Kurdish Khmer Somali
Chhattisgarhi Kikuyu South Azerbaijani
Chinese (Hans script) Kinyarwanda Southern Pashto
Chinese (Hant script) Kongo Southern Sotho
Crimean Tatar Korean Spanish
Croatian Kyrgyz Standard Arabic (Arab script)
Czech Lao Standard Arabic (Latn script)
Danish Latgalian Standard Latvian
Dutch Lingala Standard Malay
Dyula Lithuanian Swahili (individual language)
Dzongkha Luxembourgish Swati
Egyptian Arabic Macedonian Swedish
English Magahi Tajik
Estonian Maithili Tamil
Filipino Malayalam Telugu
Finnish Maltese Thai
French Manipuri Tigrinya
Galician Marathi Tosk Albanian
Ganda Minangkabau (Arab script) Uyghur
Georgian Minangkabau (Latn script) —
German Mizo —
3

Hey!

I’ve decided to try this feature but it always gets stuck while generating tts audio file. I think I set the api key correctly because I can track my token use via Google AI studio. Could you please help me about this situation?

image

1 Like
4

Fixed the issue, try again.