🎵

External Model - Text to Speech - Gemini TTS

🛠

Turn Text Into a Spoken Voiceover (Korean & English) via Google's Gemini TTS

Turn text into a natural spoken voiceover inside a CNAPS Studio flow. Six named narrator voices (Kore, Puck, Charon, Aoede, Leda, Fenrir), two model quality tiers (Flash TTS / Pro TTS), and native support for Korean and English. Runs on Google's servers via API.

🔑 Requires a Google / Gemini API key

What it does

This node connects your CNAPS Studio flows to Google's Gemini API to turn written text into spoken audio. Give it a text prompt in Korean or English, pick a narrator voice (each with a different personality) and a model tier (Flash TTS for speed and cost, Pro TTS for higher quality), and it returns a voiceover audio clip. Unlike a traditional TTS engine, Gemini TTS is a language-model-based TTS — it doesn't just pronounce words phonetically; it also understands how to say them (tone, emotion, pace) from context. This is Google's newest TTS family (Gemini 2.5 / 3.1 TTS Preview series), released as a Preview feature of the Gemini API. Unlike the on-device Qwen3-TTS node, this one runs on Google's servers — so it needs a Gemini API key and internet connectivity, but doesn't consume your local GPU.

Problem it solves

  • High-quality voiceover from text — Google's LLM-powered TTS produces natural-sounding, expressive speech that generally exceeds traditional (phoneme-based) TTS quality
  • Six distinct narrator personas — Each preset voice has a different character (firm, upbeat, informative, breezy, youthful, excitable), so you can pick one that matches your content
  • Two speed/quality tiers — Flash TTS for fast, low-cost narration; Pro TTS when quality matters most
  • Korean + English support — Bilingual out of the box (the upstream model actually supports 79 languages; the studio exposes the two most common)
  • No local GPU needed — Runs entirely on Google's infrastructure through the API
  • Podcast, audiobook, video narration — Cloud-scale audio generation for long-form scripts without local compute cost
  • Language auto-detection — In Auto mode, the model detects the language from the text itself; you can override for mixed or ambiguous content

Configuring the Node

Basic Settings

Text to speak — Input text in Korean or English.

Model variant — default Flash TTS

Variant
Best For
Flash TTS
The fast, cost-efficient tier (Gemini Flash TTS Preview)
Pro TTS
The higher-quality tier (Gemini Pro TTS Preview)

Prebuilt narrator voice — default Kore. Six named voices, each with a distinct character:

Voice
Character
Kore
Firm
Puck
Upbeat
Charon
Informative
Aoede
Breezy
Leda
Youthful
Fenrir
Excitable

Advanced Options

Spoken language hint (Auto / Korean / English, default Auto)

  • Auto — model detects the language from the input text itself (recommended)
  • Korean / English — force pronunciation in the chosen language, useful for mixed or ambiguous content

Output: Audio file (WAV) containing the spoken voiceover in the selected voice

Accuracy & Speed

Type
External API model (runs on Google's servers)
Provider
Google — Gemini API
Model variants (UI-exposed)
Flash TTS (default, fast tier) · Pro TTS (higher-quality tier)
Voices (UI-exposed)
6 named voices: Kore, Puck, Charon, Aoede, Leda, Fenrir
Voices (upstream model total)
30 — the full Gemini TTS voice library also includes Zephyr, Orus, Autonoe, Enceladus, Umbriel, Algieba, Despina, Erinome, Algenib, Rasalgethi, Laomedeia, Achernar, Alnilam, Schedar, Gacrux, Pulcherrima, Achird, Zubenelgenubi, Vindemiatrix, Sadachbia, Sadaltager, Sulafat, Iapetus, Callirrhoe — each with a different voice character
Languages (UI-exposed)
Korean, English
Languages (upstream model total)
79 — including Korean, Japanese, Chinese (Mandarin), English, Spanish, French, German, Portuguese, Arabic, Hindi, and 69 more
Requires
A valid Google / Gemini API key
Local GPU needed?
No — the studio only sends text to the API and receives audio back
Status
Preview — Gemini TTS is currently in preview status on the Gemini API

How Gemini TTS compares

Gemini TTS is the external (API-based) option in CNAPS Studio's Text-to-Speech lineup:

Gemini TTS
Qwen3-TTS (on-device)
Runs on
Google's servers (API)
Your local GPU
Requires
Gemini API key + network connectivity
Local VRAM + downloaded weights
Local GPU footprint
None
Yes
Best for
Cloud-scale narration without local compute; long-form scripts
Offline use, no API quota or per-call cost

Compliance & Provenance

Provider
Google
Provider type
GPAI — General-Purpose AI (GPAI)
License
Google Cloud Terms
EU AI Act risk class
Limited Risk
Art. 50 transparency
Required — outputs are marked. See AI Policy §2.
Region availability
Available globally
Training data summary

For more on how we classify models and mark outputs, see our AI Policy.

Run it now