Netwrck logo Netwrck

Documentation

Google speech

Gemini TTS through OpenPaths for hosted playthroughs: cast voices per character, add delivery cues, and pay per new line.

The hosted player reads dialogue with Google Gemini TTS when the reader enables it in the game menu. Nothing is generated by opening a novel, saving voice settings, or building. Each new line charges the reader's credits; replaying a line reuses cached audio at no charge.

game/speech.json

Author the cast in the editor's Playthrough voices · Google panel, or write game/speech.json by hand. Build & run includes it in the manifest. Use the displayed character names, not Ren'Py variable ids.

Cues add bracketed performance directions without changing the visible dialogue: shouting, whispers, sad, happy, excited, angry, laughing, crying, tension, gentle, serious. Unknown characters get stable automatic voices; explicit casting guarantees distinct ones.

{
  "default_voice": "Charon",
  "language": "en-US",
  "narration": false,
  "characters": {"Mara": "Kore", "Ivo": "Puck"},
  "cues": {"Get down!": "shouting", "They can hear us.": "whispers"}
}

Browser SDK

/static/js/vn/speech.js exports GOOGLE_VOICES, normalizeSpeech, speechText, speechRequest, loadSpeechSettings and installSpeech. The installed controller exposes enabled and update(request): pass the current dialogue request, or null when dialogue leaves the screen. Hosts must mute recorded voice playback while Google speech is enabled.

Advancing stops playback. An in-flight generation can still finish and be charged. The player keeps 24 clips; evicted clips may cost another call.

API and credits

POST /api/visualnovel/media/speech with input (1-1200 characters), voice and language. Same-origin requests carry the session cookie; API clients send Authorization: Bearer <secret>.

Responses carry audio_base64 or audio_url, format, credits_charged and cost_usd. Server cache hits add cached: true and charge zero credits. Cache identity covers account, host, model version, voice, language and the directed text; base64 audio persists privately within a 128 MiB budget per account and host. Failed generations are refunded and must be replayed.

curl -X POST https://netwrck.com/api/visualnovel/media/speech \
  -H 'Authorization: Bearer <secret>' -H 'Content-Type: application/json' \
  -d '{"input": "[whispers] They can hear us.", "voice": "Kore", "language": "en-US"}'

Welcome Back to Netwrck



Forgot password?

Don't have an account? Sign up here

Join Netwrck - AI Art & Chat

Create AI art, search the web, generate videos and edit photos.



Already have an account? Login here

Netwrck Unlimited

Voice Chat - Character Creation - Art Generation


Art Generator Settings


Resolution
Aesthetic Elf Styles
Anime Anime
Photographic Photographic
Digital Art Digital Art
Comic Book Comic Book
Fantasy Art Fantasy Art
Neon Punk Neon Punk
Knight Knight
Ethereal Ethereal
Low Poly Low Poly
Stunning Stunning
Line Art Line Art
Cinematic Cinematic
Wanderer Wanderer
Beggar Beggar
Seductive Seductive
Warrior Warrior
Steampunk Steampunk
Japanese Japanese
Western Western
Handsome Handsome
Pop art Pop art
Abstract Abstract
Impressionist Impressionist
Fauvist Fauvist
Beauty grace Beauty grace
Evil Evil
God God
Demon Demon
Masculine Masculine
Feminine Feminine
Majestic Majestic
Mage Mage
Princess Princess
King King
Sweet Sweet
Fantasy Fantasy

AI Generated Images


Community Images


New AI

person
This is what your AI will say when someone first chats with them
Help others understand what your AI character is about

AI Settings

Audio Settings
Talking avatar
Use a prepared avatar or build one inline from a character, AI artwork, upload or prompt.
Character calls

Local is free (browser speech + your chat model + on-device TTS). Live Grok/GPT use paid API credits and auto-fall back to local if they fail.

1.0x

Translation

For Chinese: hanzi + pinyin per word, tap any word to hear it spoken locally.


Art Generation

When enabled, AI can automatically generate art based on conversations

Prepended to art prompts when Make Art runs.
Only applies when GPT Image 2 is selected. GPT Image 2 chat art consumes credits.

High Resolution

LTX 2.3 Image to Video

Selected image
Pick any generated image and turn it into a video.
$0.58 · 58 credits
Your latest request will appear here.

Saved Requests

Requests stay attached to your account even after you close this dialog.
No video requests yet.

Find an AI to chat with

No AI characters found.

Help Everyone Know You

person

Share Image

Consider our affiliate program to earn!

Copy

Reddit

X

Facebook

LinkedIn

NETW Coin is live!

AI should be owned by everyone.
So we are building a new AI economy together.
Thank you for being on this journey with us!
Read More

Netwrck Android App

Our Android App Is Available Now - Please Check it out!

New Post

Add Netwrck Credits

Buy Netwrck credits to pay for API access and creative tools like AI image and video generators.

Current Balance: $0.00
Custom:

Sign in to keep chatting

You have used your messages. Sign in to continue this conversation and keep your chat history.

No card required.

Characters in this room

Remove a current character, add an existing one, or create a new character for this room.

Add an existing character