Speech synthesis and narration

File: config/system/system-speech.json. Side: client. A missing file is read after being copied from the file of the same name under the JAR's defaults.

Field	Type	Required	Default	Description
enabled	Boolean	no	false	Enables narration; information such as the API key must still be valid.
provider	String	no	"aliyun"	Provider ID; aliyun is built in.
apiKey	String	no	""	Personal Bailian API key, kept only on the client.
region	String	no	"beijing"	Service region; must match the key / voice.
workspace	String	no	""	Optional Workspace ID.
model	String	no	"qwen3-tts-flash"	Synthesis model, qwen3-tts-flash by default.
voice	String	no	"Cherry"	Default or custom voice ID; must match the model.
language	String	no	"Client"	Client follows the client language, or a synthesis language supported by the service, such as Auto, Chinese or English.
volume	Integer	no	80	Narration volume, 0–100.
cacheEnabled	Boolean	no	false	Local audio cache, off by default.
questRequirements	Boolean	no	true	Narrate quest requirements.
questRewards	Boolean	no	true	Narrate quest rewards.
questDescription	Boolean	no	false	Narrate the filtered quest background, off by default.
notificationSoundEnabled	Boolean	no	true	Sound when a notification is first expanded, at 75% of the narration volume.

These credentials and the speech usage belong only to the current client and are not synced to the server. Prefer editing them in the speech tab of /system settings; settings are marked and saved automatically after being submitted. Closing a notification, confirming early or switching the speech session cancels the corresponding narration.

Bundled default file

config/system/system-speech.json

{
  "enabled": false,
  "provider": "aliyun",
  "apiKey": "",
  "region": "beijing",
  "workspace": "",
  "model": "qwen3-tts-flash",
  "voice": "Cherry",
  "language": "Client",
  "volume": 80,
  "cacheEnabled": false,
  "questRequirements": true,
  "questRewards": true,
  "questDescription": false,
  "notificationSoundEnabled": true
}