kugelaudio 0.1.9 → 0.1.11

package/README.md

@@ -31,7 +31,7 @@ const client = new KugelAudio({ apiKey: 'your_api_key' });
 // Generate speech
 const audio = await client.tts.generate({
   text: 'Hello, world!',
-  model: 'kugel-1-turbo',
+  modelId: 'kugel-1-turbo',
 });
 
 // Create a playable blob (browser)
@@ -144,12 +144,11 @@ Generate complete audio and receive it all at once:
 ```typescript
 const audio = await client.tts.generate({
   text: 'Hello, this is a test of the KugelAudio text-to-speech system.',
-  model: 'kugel-1-turbo',  // 'kugel-1-turbo' (fast) or 'kugel-1' (quality)
+  modelId: 'kugel-1-turbo',  // 'kugel-1-turbo' (fast) or 'kugel-1' (quality)
   voiceId: 123,              // Optional: specific voice ID
   cfgScale: 2.0,             // Guidance scale (1.0-5.0)
   maxNewTokens: 2048,        // Maximum tokens to generate
   sampleRate: 24000,         // Output sample rate
-  speakerPrefix: true,       // Add speaker prefix for better quality
   normalize: true,           // Enable text normalization (see below)
   language: 'en',            // Language for normalization
 });
@@ -171,7 +170,7 @@ import { createWavBlob } from 'kugelaudio';
 
 const audio = await client.tts.generate({
   text: 'Hello, world!',
-  model: 'kugel-1-turbo',
+  modelId: 'kugel-1-turbo',
 });
 
 // Create WAV blob for playback
@@ -200,7 +199,7 @@ Receive audio chunks as they are generated for lower latency:
 await client.tts.stream(
   {
     text: 'Hello, this is streaming audio.',
-    model: 'kugel-1-turbo',
+    modelId: 'kugel-1-turbo',
   },
   {
     onOpen: () => {
@@ -354,8 +353,7 @@ interface GenerateOptions {
   cfgScale?: number;       // Default: 2.0
   maxNewTokens?: number;   // Default: 2048
   sampleRate?: number;     // Default: 24000
-  speakerPrefix?: boolean; // Default: true
-  normalize?: boolean;     // Default: false - Enable text normalization
+  normalize?: boolean;     // Default: true - Enable text normalization
   language?: string;       // ISO 639-1 code for normalization (e.g., 'en', 'de')
 }
 ```
@@ -520,7 +518,7 @@ async function main() {
   await client.tts.stream(
     {
       text: 'Welcome to KugelAudio. This is an example of high-quality text-to-speech synthesis.',
-      model: 'kugel-1-turbo',
+      modelId: 'kugel-1-turbo',
     },
     {
       onChunk: (chunk) => {

package/dist/index.d.mts

@@ -15,7 +15,7 @@ interface Model {
 /**
  * Voice category types.
  */
-type VoiceCategory = 'premade' | 'cloned' | 'designed';
+type VoiceCategory = 'premade' | 'cloned' | 'designed' | 'conversational' | 'narrative' | 'narrative_story' | 'characters';
 /**
  * Voice sex types.
  */
@@ -48,7 +48,7 @@ interface GenerateOptions {
     /** Text to synthesize */
     text: string;
     /** Model to use: 'kugel-1-turbo' (1.5B, fast) or 'kugel-1' (7B, premium). Default: 'kugel-1-turbo' */
-    model?: string;
+    modelId?: string;
     /** Voice ID to use */
     voiceId?: number;
     /** CFG scale for generation (default: 2.0) */
@@ -57,21 +57,18 @@ interface GenerateOptions {
     maxNewTokens?: number;
     /** Output sample rate (default: 24000) */
     sampleRate?: number;
-    /** Whether to add speaker prefix (default: true) */
-    speakerPrefix?: boolean;
     /**
      * Enable text normalization (converts numbers, dates, etc. to spoken words).
      * When true, text will be normalized before TTS generation.
-     * Default: false
+     * Default: true
      *
-     * ⚠️ WARNING: Using normalize=true without specifying language adds ~150ms
-     * latency for language auto-detection. For best performance, always specify
-     * the language parameter when using normalization.
+     * ⚠️ For best performance, always specify the language parameter when using
+     * normalization. Without it, language auto-detection adds ~150ms latency.
      */
     normalize?: boolean;
     /**
      * ISO 639-1 language code for text normalization (e.g., 'de', 'en', 'fr').
-     * If not provided and normalize is true, language will be auto-detected
+     * If not provided and normalize is true (default), language will be auto-detected
      * (adds ~150ms latency).
      *
      * Supported: de, en, fr, es, it, pt, nl, pl, sv, da, no, fi, cs, hu, ro,
@@ -91,12 +88,20 @@ interface StreamConfig {
     maxNewTokens?: number;
     /** Output sample rate */
     sampleRate?: number;
-    /** Whether to add speaker prefix */
-    speakerPrefix?: boolean;
     /** Auto-flush timeout in milliseconds */
     flushTimeoutMs?: number;
     /** Maximum buffer length */
     maxBufferLength?: number;
+    /**
+     * Enable text normalization (converts numbers, dates, etc. to spoken words).
+     * Default: true
+     */
+    normalize?: boolean;
+    /**
+     * ISO 639-1 language code for text normalization (e.g., 'de', 'en', 'fr').
+     * Specify to avoid ~150ms auto-detection latency.
+     */
+    language?: string;
 }
 /**
  * Audio chunk from streaming TTS.
@@ -180,7 +185,7 @@ interface KugelAudioOptions {
     orgId?: number;
     /** API base URL (default: https://api.kugelaudio.com) */
     apiUrl?: string;
-    /** TTS server URL (default: https://eu.kugelaudio.com) */
+    /** TTS server URL (default: same as apiUrl) */
     ttsUrl?: string;
     /** Request timeout in milliseconds (default: 60000) */
     timeout?: number;
@@ -199,8 +204,6 @@ interface MultiContextConfig {
     maxNewTokens?: number;
     /** Enable text normalization (default: true) */
     normalize?: boolean;
-    /** Add speaker prefix (default: true) */
-    speakerPrefix?: boolean;
     /** Seconds before context auto-closes (default: 20.0) */
     inactivityTimeout?: number;
 }
@@ -459,13 +462,13 @@ declare class MultiContextSession {
  * // Generate audio with fast model (1.5B params)
  * const audio = await client.tts.generate({
  *   text: 'Hello, world!',
- *   model: 'kugel-1-turbo',
+ *   modelId: 'kugel-1-turbo',
  * });
  *
  * // Generate audio with premium model (7B params)
  * const audio = await client.tts.generate({
  *   text: 'Hello, world!',
- *   model: 'kugel-1',
+ *   modelId: 'kugel-1',
  * });
  * ```
  */

package/dist/index.d.ts

@@ -15,7 +15,7 @@ interface Model {
 /**
  * Voice category types.
  */
-type VoiceCategory = 'premade' | 'cloned' | 'designed';
+type VoiceCategory = 'premade' | 'cloned' | 'designed' | 'conversational' | 'narrative' | 'narrative_story' | 'characters';
 /**
  * Voice sex types.
  */
@@ -48,7 +48,7 @@ interface GenerateOptions {
     /** Text to synthesize */
     text: string;
     /** Model to use: 'kugel-1-turbo' (1.5B, fast) or 'kugel-1' (7B, premium). Default: 'kugel-1-turbo' */
-    model?: string;
+    modelId?: string;
     /** Voice ID to use */
     voiceId?: number;
     /** CFG scale for generation (default: 2.0) */
@@ -57,21 +57,18 @@ interface GenerateOptions {
     maxNewTokens?: number;
     /** Output sample rate (default: 24000) */
     sampleRate?: number;
-    /** Whether to add speaker prefix (default: true) */
-    speakerPrefix?: boolean;
     /**
      * Enable text normalization (converts numbers, dates, etc. to spoken words).
      * When true, text will be normalized before TTS generation.
-     * Default: false
+     * Default: true
      *
-     * ⚠️ WARNING: Using normalize=true without specifying language adds ~150ms
-     * latency for language auto-detection. For best performance, always specify
-     * the language parameter when using normalization.
+     * ⚠️ For best performance, always specify the language parameter when using
+     * normalization. Without it, language auto-detection adds ~150ms latency.
      */
     normalize?: boolean;
     /**
      * ISO 639-1 language code for text normalization (e.g., 'de', 'en', 'fr').
-     * If not provided and normalize is true, language will be auto-detected
+     * If not provided and normalize is true (default), language will be auto-detected
      * (adds ~150ms latency).
      *
      * Supported: de, en, fr, es, it, pt, nl, pl, sv, da, no, fi, cs, hu, ro,
@@ -91,12 +88,20 @@ interface StreamConfig {
     maxNewTokens?: number;
     /** Output sample rate */
     sampleRate?: number;
-    /** Whether to add speaker prefix */
-    speakerPrefix?: boolean;
     /** Auto-flush timeout in milliseconds */
     flushTimeoutMs?: number;
     /** Maximum buffer length */
     maxBufferLength?: number;
+    /**
+     * Enable text normalization (converts numbers, dates, etc. to spoken words).
+     * Default: true
+     */
+    normalize?: boolean;
+    /**
+     * ISO 639-1 language code for text normalization (e.g., 'de', 'en', 'fr').
+     * Specify to avoid ~150ms auto-detection latency.
+     */
+    language?: string;
 }
 /**
  * Audio chunk from streaming TTS.
@@ -180,7 +185,7 @@ interface KugelAudioOptions {
     orgId?: number;
     /** API base URL (default: https://api.kugelaudio.com) */
     apiUrl?: string;
-    /** TTS server URL (default: https://eu.kugelaudio.com) */
+    /** TTS server URL (default: same as apiUrl) */
     ttsUrl?: string;
     /** Request timeout in milliseconds (default: 60000) */
     timeout?: number;
@@ -199,8 +204,6 @@ interface MultiContextConfig {
     maxNewTokens?: number;
     /** Enable text normalization (default: true) */
     normalize?: boolean;
-    /** Add speaker prefix (default: true) */
-    speakerPrefix?: boolean;
     /** Seconds before context auto-closes (default: 20.0) */
     inactivityTimeout?: number;
 }
@@ -459,13 +462,13 @@ declare class MultiContextSession {
  * // Generate audio with fast model (1.5B params)
  * const audio = await client.tts.generate({
  *   text: 'Hello, world!',
- *   model: 'kugel-1-turbo',
+ *   modelId: 'kugel-1-turbo',
  * });
  *
  * // Generate audio with premium model (7B params)
  * const audio = await client.tts.generate({
  *   text: 'Hello, world!',
- *   model: 'kugel-1',
+ *   modelId: 'kugel-1',
  * });
  * ```
  */

package/dist/index.js

@@ -415,13 +415,12 @@ var TTSResource = class {
       callbacks.onOpen?.();
       ws.send(JSON.stringify({
         text: options.text,
-        model: options.model || "kugel-1-turbo",
+        model_id: options.modelId || "kugel-1-turbo",
         voice_id: options.voiceId,
         cfg_scale: options.cfgScale ?? 2,
         max_new_tokens: options.maxNewTokens ?? 2048,
         sample_rate: options.sampleRate ?? 24e3,
-        speaker_prefix: options.speakerPrefix ?? true,
-        normalize: options.normalize ?? false,
+        normalize: options.normalize ?? true,
         ...options.language && { language: options.language }
       }));
     });
@@ -437,13 +436,12 @@ var TTSResource = class {
         callbacks.onOpen?.();
         ws.send(JSON.stringify({
           text: options.text,
-          model: options.model || "kugel-1-turbo",
+          model_id: options.modelId || "kugel-1-turbo",
           voice_id: options.voiceId,
           cfg_scale: options.cfgScale ?? 2,
           max_new_tokens: options.maxNewTokens ?? 2048,
           sample_rate: options.sampleRate ?? 24e3,
-          speaker_prefix: options.speakerPrefix ?? true,
-          normalize: options.normalize ?? false,
+          normalize: options.normalize ?? true,
           ...options.language && { language: options.language }
         }));
       };
@@ -676,7 +674,6 @@ var MultiContextSession = class {
       if (this.config.cfgScale) msg.cfg_scale = this.config.cfgScale;
       if (this.config.maxNewTokens) msg.max_new_tokens = this.config.maxNewTokens;
       if (this.config.normalize !== void 0) msg.normalize = this.config.normalize;
-      if (this.config.speakerPrefix !== void 0) msg.speaker_prefix = this.config.speakerPrefix;
       if (this.config.inactivityTimeout) msg.inactivity_timeout = this.config.inactivityTimeout;
     }
     const voiceId = options?.voiceId || this.config.defaultVoiceId;

package/dist/index.mjs

@@ -379,13 +379,12 @@ var TTSResource = class {
       callbacks.onOpen?.();
       ws.send(JSON.stringify({
         text: options.text,
-        model: options.model || "kugel-1-turbo",
+        model_id: options.modelId || "kugel-1-turbo",
         voice_id: options.voiceId,
         cfg_scale: options.cfgScale ?? 2,
         max_new_tokens: options.maxNewTokens ?? 2048,
         sample_rate: options.sampleRate ?? 24e3,
-        speaker_prefix: options.speakerPrefix ?? true,
-        normalize: options.normalize ?? false,
+        normalize: options.normalize ?? true,
         ...options.language && { language: options.language }
       }));
     });
@@ -401,13 +400,12 @@ var TTSResource = class {
         callbacks.onOpen?.();
         ws.send(JSON.stringify({
           text: options.text,
-          model: options.model || "kugel-1-turbo",
+          model_id: options.modelId || "kugel-1-turbo",
           voice_id: options.voiceId,
           cfg_scale: options.cfgScale ?? 2,
           max_new_tokens: options.maxNewTokens ?? 2048,
           sample_rate: options.sampleRate ?? 24e3,
-          speaker_prefix: options.speakerPrefix ?? true,
-          normalize: options.normalize ?? false,
+          normalize: options.normalize ?? true,
           ...options.language && { language: options.language }
         }));
       };
@@ -640,7 +638,6 @@ var MultiContextSession = class {
       if (this.config.cfgScale) msg.cfg_scale = this.config.cfgScale;
       if (this.config.maxNewTokens) msg.max_new_tokens = this.config.maxNewTokens;
       if (this.config.normalize !== void 0) msg.normalize = this.config.normalize;
-      if (this.config.speakerPrefix !== void 0) msg.speaker_prefix = this.config.speakerPrefix;
       if (this.config.inactivityTimeout) msg.inactivity_timeout = this.config.inactivityTimeout;
     }
     const voiceId = options?.voiceId || this.config.defaultVoiceId;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "kugelaudio",
-  "version": "0.1.9",
+  "version": "0.1.11",
   "description": "Official JavaScript/TypeScript SDK for KugelAudio TTS API",
   "main": "dist/index.js",
   "module": "dist/index.mjs",

package/src/client.ts

@@ -369,13 +369,12 @@ class TTSResource {
 
       ws.send(JSON.stringify({
         text: options.text,
-        model: options.model || 'kugel-1-turbo',
+        model_id: options.modelId || 'kugel-1-turbo',
         voice_id: options.voiceId,
         cfg_scale: options.cfgScale ?? 2.0,
         max_new_tokens: options.maxNewTokens ?? 2048,
         sample_rate: options.sampleRate ?? 24000,
-        speaker_prefix: options.speakerPrefix ?? true,
-        normalize: options.normalize ?? false,
+        normalize: options.normalize ?? true,
         ...(options.language && { language: options.language }),
       }));
     });
@@ -397,13 +396,12 @@ class TTSResource {
         // Send TTS request
         ws.send(JSON.stringify({
           text: options.text,
-          model: options.model || 'kugel-1-turbo',
+          model_id: options.modelId || 'kugel-1-turbo',
           voice_id: options.voiceId,
           cfg_scale: options.cfgScale ?? 2.0,
           max_new_tokens: options.maxNewTokens ?? 2048,
           sample_rate: options.sampleRate ?? 24000,
-          speaker_prefix: options.speakerPrefix ?? true,
-          normalize: options.normalize ?? false,
+          normalize: options.normalize ?? true,
           ...(options.language && { language: options.language }),
         }));
       };
@@ -686,7 +684,6 @@ class MultiContextSession {
       if (this.config.cfgScale) msg.cfg_scale = this.config.cfgScale;
       if (this.config.maxNewTokens) msg.max_new_tokens = this.config.maxNewTokens;
       if (this.config.normalize !== undefined) msg.normalize = this.config.normalize;
-      if (this.config.speakerPrefix !== undefined) msg.speaker_prefix = this.config.speakerPrefix;
       if (this.config.inactivityTimeout) msg.inactivity_timeout = this.config.inactivityTimeout;
     }
 
@@ -807,13 +804,13 @@ class MultiContextSession {
  * // Generate audio with fast model (1.5B params)
  * const audio = await client.tts.generate({
  *   text: 'Hello, world!',
- *   model: 'kugel-1-turbo',
+ *   modelId: 'kugel-1-turbo',
  * });
  *
  * // Generate audio with premium model (7B params)
  * const audio = await client.tts.generate({
  *   text: 'Hello, world!',
- *   model: 'kugel-1',
+ *   modelId: 'kugel-1',
  * });
  * ```
  */

package/src/index.ts

@@ -18,13 +18,13 @@
  * // Generate audio (non-streaming)
  * const audio = await client.tts.generate({
  *   text: 'Hello, world!',
- *   model: 'kugel-1-turbo',
+ *   modelId: 'kugel-1-turbo',
  *   voiceId: 123,
  * });
  *
  * // Generate audio (streaming)
  * await client.tts.stream(
- *   { text: 'Hello, world!', model: 'kugel-1-turbo' },
+ *   { text: 'Hello, world!', modelId: 'kugel-1-turbo' },
  *   {
  *     onChunk: (chunk) => {
  *       // Process audio chunk

package/src/types.ts

@@ -17,7 +17,7 @@ export interface Model {
 /**
  * Voice category types.
  */
-export type VoiceCategory = 'premade' | 'cloned' | 'designed';
+export type VoiceCategory = 'premade' | 'cloned' | 'designed' | 'conversational' | 'narrative' | 'narrative_story' | 'characters';
 
 /**
  * Voice sex types.
@@ -54,7 +54,7 @@ export interface GenerateOptions {
   /** Text to synthesize */
   text: string;
   /** Model to use: 'kugel-1-turbo' (1.5B, fast) or 'kugel-1' (7B, premium). Default: 'kugel-1-turbo' */
-  model?: string;
+  modelId?: string;
   /** Voice ID to use */
   voiceId?: number;
   /** CFG scale for generation (default: 2.0) */
@@ -63,21 +63,18 @@ export interface GenerateOptions {
   maxNewTokens?: number;
   /** Output sample rate (default: 24000) */
   sampleRate?: number;
-  /** Whether to add speaker prefix (default: true) */
-  speakerPrefix?: boolean;
   /** 
    * Enable text normalization (converts numbers, dates, etc. to spoken words).
    * When true, text will be normalized before TTS generation.
-   * Default: false
+   * Default: true
    * 
-   * ⚠️ WARNING: Using normalize=true without specifying language adds ~150ms
-   * latency for language auto-detection. For best performance, always specify
-   * the language parameter when using normalization.
+   * ⚠️ For best performance, always specify the language parameter when using
+   * normalization. Without it, language auto-detection adds ~150ms latency.
    */
   normalize?: boolean;
   /**
    * ISO 639-1 language code for text normalization (e.g., 'de', 'en', 'fr').
-   * If not provided and normalize is true, language will be auto-detected
+   * If not provided and normalize is true (default), language will be auto-detected
    * (adds ~150ms latency).
    * 
    * Supported: de, en, fr, es, it, pt, nl, pl, sv, da, no, fi, cs, hu, ro,
@@ -98,12 +95,20 @@ export interface StreamConfig {
   maxNewTokens?: number;
   /** Output sample rate */
   sampleRate?: number;
-  /** Whether to add speaker prefix */
-  speakerPrefix?: boolean;
   /** Auto-flush timeout in milliseconds */
   flushTimeoutMs?: number;
   /** Maximum buffer length */
   maxBufferLength?: number;
+  /**
+   * Enable text normalization (converts numbers, dates, etc. to spoken words).
+   * Default: true
+   */
+  normalize?: boolean;
+  /**
+   * ISO 639-1 language code for text normalization (e.g., 'de', 'en', 'fr').
+   * Specify to avoid ~150ms auto-detection latency.
+   */
+  language?: string;
 }
 
 /**
@@ -192,7 +197,7 @@ export interface KugelAudioOptions {
   orgId?: number;
   /** API base URL (default: https://api.kugelaudio.com) */
   apiUrl?: string;
-  /** TTS server URL (default: https://eu.kugelaudio.com) */
+  /** TTS server URL (default: same as apiUrl) */
   ttsUrl?: string;
   /** Request timeout in milliseconds (default: 60000) */
   timeout?: number;
@@ -221,8 +226,6 @@ export interface MultiContextConfig {
   maxNewTokens?: number;
   /** Enable text normalization (default: true) */
   normalize?: boolean;
-  /** Add speaker prefix (default: true) */
-  speakerPrefix?: boolean;
   /** Seconds before context auto-closes (default: 20.0) */
   inactivityTimeout?: number;
 }