Interface TTSV3Options

Options specific to bulbul:v3

interface TTSV3Options {
    apiKey?: string;
    baseURL?: string;
    model: "bulbul:v3";
    pace?: number;
    sampleRate?: number;
    sentenceTokenizer?: tokenize.SentenceTokenizer;
    speaker?: string;
    streaming?: boolean;
    targetLanguageCode?: string;
    temperature?: number;
}

Hierarchy

  • TTSBaseOptions
    • TTSV3Options

Properties

apiKey? baseURL? model pace? sampleRate? sentenceTokenizer? speaker? streaming? targetLanguageCode? temperature?

Properties

Optional apiKey

apiKey?: string

Sarvam API key. Defaults to $SARVAM_API_KEY

Optional baseURL

baseURL?: string

Base URL for the Sarvam API

model

model: "bulbul:v3"

Optional pace

pace?: number

Speech pace. v2: 0.3–3.0, v3: 0.5–2.0 (default 1.0)

Optional sampleRate

sampleRate?: number

Output sample rate in Hz (default 24000)

Optional sentenceTokenizer

sentenceTokenizer?: tokenize.SentenceTokenizer

Sentence tokenizer for streaming (default: basic sentence tokenizer)

Optional speaker

speaker?: string

Speaker voice (v3 voices). Default: 'shubh'

Optional streaming

streaming?: boolean

Whether to use native WebSocket streaming for stream(). Set to false to prefer non-streaming REST synthesis (used by Agent via TTS StreamAdapter). Default: true.

Optional targetLanguageCode

targetLanguageCode?: string

Target language code (BCP-47)

Optional temperature

temperature?: number

Temperature for voice variation, 0.01 to 2.0 (v3 only, default 0.6)

Settings

Member Visibility

Theme

On This Page

apiKeybaseURLmodelpacesampleRatesentenceTokenizerspeakerstreamingtargetLanguageCodetemperature