playkit-sdk 1.2.13 → 1.4.0-beta.1

package/dist/playkit-sdk.d.ts

@@ -413,6 +413,8 @@ interface SDKConfig {
     defaultImageModel?: string;
     /** Default transcription model to use */
     defaultTranscriptionModel?: string;
+    /** Default text-to-speech model to use */
+    defaultTTSModel?: string;
     /**
      * Enable debug logging
      * @deprecated Use `logging.level` instead. Will be removed in v2.0.
@@ -657,10 +659,12 @@ interface ChatCompletionResponse {
  * Streaming chunk formats
  */
 interface StreamChunk {
-    type: 'text-delta' | 'done' | 'error';
+    type: 'text-delta' | 'done' | 'finish' | 'abort' | 'error';
     id?: string;
     delta?: string;
     error?: string;
+    errorText?: string;
+    reason?: string;
 }
 /**
  * NPC Action parameter types
@@ -874,6 +878,85 @@ interface TranscriptionResponse {
     segments?: TranscriptionSegment[];
 }
 
+/**
+ * Text-to-speech (TTS) type definitions
+ */
+/**
+ * Configuration for text-to-speech requests
+ */
+interface TTSConfig {
+    /**
+     * Text to synthesize into speech (max 10000 characters)
+     */
+    text: string;
+    /**
+     * Model to use for synthesis
+     * Defaults to 'default-tts-model' (alias resolved by the backend)
+     */
+    model?: string;
+    /**
+     * Voice id to use (e.g., 'male-qn-qingse')
+     */
+    voice?: string;
+    /**
+     * Playback speed multiplier
+     */
+    speed?: number;
+    /**
+     * Volume
+     */
+    vol?: number;
+    /**
+     * Pitch adjustment
+     */
+    pitch?: number;
+    /**
+     * Emotion of the speech (e.g., 'happy', 'sad')
+     */
+    emotion?: string;
+    /**
+     * Output audio format (e.g., 'mp3', 'wav')
+     */
+    format?: string;
+    /**
+     * Language boost hint to improve pronunciation for a specific language
+     */
+    languageBoost?: string;
+    /**
+     * Passthrough voice settings object for advanced configuration
+     */
+    voiceSetting?: Record<string, unknown>;
+    /**
+     * Passthrough audio settings object for advanced configuration
+     */
+    audioSetting?: Record<string, unknown>;
+}
+/**
+ * Options for simplified text-to-speech methods (everything except the text)
+ */
+type TTSOptions = Omit<TTSConfig, 'text'>;
+/**
+ * Result of a text-to-speech request
+ */
+interface TTSResult {
+    /**
+     * Raw audio bytes
+     */
+    audio: ArrayBuffer;
+    /**
+     * Audio format / content type (e.g., 'audio/mpeg' or 'mp3')
+     */
+    format: string;
+    /**
+     * Number of characters billed for this request
+     */
+    usageCharacters: number;
+    /**
+     * Length of the generated audio in milliseconds (if reported)
+     */
+    audioLengthMs?: number;
+}
+
 /**
  * Device Authorization Flow Manager
  * Manages Device Auth polling flow for desktop/CLI/Unity applications
@@ -1261,7 +1344,7 @@ interface RechargeConfig {
     checkBalanceAfterApiCall?: boolean;
     /**
      * Base URL for the recharge portal
-     * @default 'https://playkit.ai/recharge'
+     * @default 'https://players.playkit.ai/recharge'
      */
     rechargePortalUrl?: string;
     /**
@@ -1909,6 +1992,59 @@ declare class TranscriptionClient {
     transcribeFile(file: File, options?: TranscriptionOptions): Promise<TranscriptionResult>;
 }
 
+/**
+ * TTS provider for HTTP communication with the text-to-speech API
+ */
+
+declare class TTSProvider {
+    private authManager;
+    private config;
+    private baseURL;
+    private playerClient?;
+    constructor(authManager: AuthManager, config: SDKConfig);
+    /**
+     * Set player client for balance checking
+     */
+    setPlayerClient(playerClient: PlayerClient): void;
+    /**
+     * Synthesize text into speech audio
+     */
+    synthesize(ttsConfig: TTSConfig): Promise<TTSResult>;
+}
+
+/**
+ * High-level client for text-to-speech synthesis
+ */
+
+declare class TTSClient {
+    private provider;
+    private model;
+    constructor(provider: TTSProvider, model?: string);
+    /**
+     * Get the current model name
+     */
+    get modelName(): string;
+    /**
+     * Synthesize text into speech audio
+     * @param config - Full TTS configuration
+     * @returns TTS result containing raw audio bytes and usage metadata
+     */
+    synthesize(config: TTSConfig): Promise<TTSResult>;
+    /**
+     * Synthesize text into speech and return it as a Blob (browser-friendly)
+     * @param config - Full TTS configuration
+     * @returns Audio Blob with the appropriate MIME type
+     */
+    synthesizeToBlob(config: TTSConfig): Promise<Blob>;
+    /**
+     * Synthesize text into speech and return an object URL (browser only)
+     * @param config - Full TTS configuration
+     * @returns An object URL that can be assigned to an <audio> element
+     * @throws PlayKitError if URL.createObjectURL is unavailable (e.g. Node.js)
+     */
+    synthesizeToObjectURL(config: TTSConfig): Promise<string>;
+}
+
 /**
  * NPC Client for simplified conversation management
  * Automatically handles conversation history
@@ -2034,10 +2170,11 @@ declare class NPCClient extends EventEmitter {
     /**
      * Manually generate reply predictions based on current conversation.
      * Uses the fast model for quick generation.
+     * @param tempPrompt Optional temporary prompt to influence the prediction style/tone
      * @param count Number of predictions to generate (default: uses predictionCount property)
      * @returns Array of predicted player replies, or empty array on failure
      */
-    generateReplyPredictions(count?: number): Promise<string[]>;
+    generateReplyPredictions(tempPrompt?: string, count?: number): Promise<string[]>;
     /**
      * Parse predictions from JSON array response
      */
@@ -2195,6 +2332,8 @@ declare class AIContextManager extends EventEmitter<AIContextManagerEvents> {
     private static _instance;
     private config;
     private playerDescription;
+    private playerPrompt;
+    private playerMemories;
     private npcStates;
     private autoCompactTimer;
     private chatClientFactory;
@@ -2233,6 +2372,47 @@ declare class AIContextManager extends EventEmitter<AIContextManagerEvents> {
      * Clear the player description.
      */
     clearPlayerDescription(): void;
+    /**
+     * Set the player's character prompt/persona.
+     * This defines how the player character speaks and behaves.
+     * Used when generating reply predictions to match the player's tone.
+     * @param prompt The player character's persona/prompt
+     */
+    setPlayerPrompt(prompt: string | null): void;
+    /**
+     * Get the current player prompt.
+     * @returns The player prompt, or null if not set
+     */
+    getPlayerPrompt(): string | null;
+    /**
+     * Set or update a memory for the player character.
+     * Memories are appended to the player prompt to form the full player context.
+     * Set memoryContent to null or empty to remove the memory.
+     * @param memoryName The name/key of the memory
+     * @param memoryContent The content of the memory. Null or empty to remove.
+     */
+    setPlayerMemory(memoryName: string, memoryContent: string | null): void;
+    /**
+     * Get a specific player memory by name.
+     * @param memoryName The name of the memory to retrieve
+     * @returns The memory content, or undefined if not found
+     */
+    getPlayerMemory(memoryName: string): string | undefined;
+    /**
+     * Get all player memory names currently stored.
+     * @returns Array of memory names
+     */
+    getPlayerMemoryNames(): string[];
+    /**
+     * Clear all player memories (but keep player prompt).
+     */
+    clearPlayerMemories(): void;
+    /**
+     * Build the complete player context from PlayerPrompt + PlayerMemories.
+     * Used by NPCClient for generating reply predictions.
+     * @returns The combined player context string, or null if no context is set
+     */
+    buildPlayerContext(): string | null;
     /**
      * Register an NPC for context management.
      * @param npc The NPC client to register
@@ -2313,6 +2493,7 @@ declare class PlayKitSDK extends EventEmitter {
     private chatProvider;
     private imageProvider;
     private transcriptionProvider;
+    private ttsProvider;
     private contextManager;
     private schemaLibrary;
     private initialized;
@@ -2382,6 +2563,11 @@ declare class PlayKitSDK extends EventEmitter {
      * @param model - Transcription model to use (default: 'whisper-large')
      */
     createTranscriptionClient(model?: string): TranscriptionClient;
+    /**
+     * Create a TTS client for text-to-speech
+     * @param model - TTS model to use (default: 'default-tts-model')
+     */
+    createTTSClient(model?: string): TTSClient;
     /**
      * Create an NPC client
      * Automatically registers with AIContextManager
@@ -2682,8 +2868,8 @@ declare class TokenStorage {
 declare class AuthFlowManager extends EventEmitter {
     private baseURL;
     private currentSessionId;
-    private uiContainer;
-    private isSuccess;
+    private _uiContainer;
+    private _isSuccess;
     private currentLanguage;
     private modal;
     private identifierPanel;
@@ -2903,5 +3089,13 @@ declare class TokenValidator {
  */
 declare const defaultTokenValidator: TokenValidator;
 
-export { AIContextManager, AuthFlowManager, AuthManager, BrowserStorage, BufferLogHandler, CallbackLogHandler, ChatClient, DeviceAuthFlowManager, ImageClient, LogLevel, Logger, MemoryStorage, NPCClient, PlayKitError, PlayKitSDK, PlayerClient, RechargeManager, SchemaLibrary, StreamParser, TokenStorage, TokenValidator, TranscriptionClient, createMultimodalMessage, createStorage, createTextMessage, PlayKitSDK as default, defaultContextManager, defaultSchemaLibrary, defaultTokenValidator, isLocalStorageAvailable };
-export type { AIContextManagerConfig, AIContextManagerEvents, APIResult, AudioContentPart, AuthState, ChatCompletionResponse, ChatConfig, ChatResult, ChatStreamConfig, ChatWithToolsConfig, ChatWithToolsStreamConfig, ConversationSaveData, DeveloperTokenFallbackConfig, DeviceAuthFlowOptions, DeviceAuthInitResult, DeviceAuthResult, GameInfo, GeneratedImage, IStorage, ImageContentPart, ImageGenerationConfig, ImageGenerationResponse, ImageInput, ImageSize, LogConfig, LogEntry, LogHandler, MemoryEntry, Message, MessageContent, MessageContentPart, MessageRole, NPCConfig, PlayerInfo, RechargeConfig, RechargeEvents, RechargeModalOptions, SDKConfig, SDKMode, SchemaEntry, SetNicknameRequest, SetNicknameResponse, StreamChunk, StructuredGenerationConfig, StructuredOutputConfig, StructuredResult, TextContentPart, TokenRefreshResult, TokenScope, TokenStorageOptions, TokenValidatorOptions, TokenVerificationResult, TranscriptionConfig, TranscriptionOptions, TranscriptionResult, TranscriptionSegment, ValidatedPlayerInfo };
+declare global {
+    interface Window {
+        PlayKitSDK: typeof PlayKitSDK & {
+            PlayKitSDK: typeof PlayKitSDK;
+        };
+    }
+}
+
+export { AIContextManager, AuthFlowManager, AuthManager, BrowserStorage, BufferLogHandler, CallbackLogHandler, ChatClient, DeviceAuthFlowManager, ImageClient, LogLevel, Logger, MemoryStorage, NPCClient, PlayKitError, PlayKitSDK, PlayerClient, RechargeManager, SchemaLibrary, StreamParser, TTSClient, TokenStorage, TokenValidator, TranscriptionClient, createMultimodalMessage, createStorage, createTextMessage, PlayKitSDK as default, defaultContextManager, defaultSchemaLibrary, defaultTokenValidator, isLocalStorageAvailable };
+export type { AIContextManagerConfig, AIContextManagerEvents, APIResult, AudioContentPart, AuthState, ChatCompletionResponse, ChatConfig, ChatResult, ChatStreamConfig, ChatWithToolsConfig, ChatWithToolsStreamConfig, ConversationSaveData, DeveloperTokenFallbackConfig, DeviceAuthFlowOptions, DeviceAuthInitResult, DeviceAuthResult, GameInfo, GeneratedImage, IStorage, ImageContentPart, ImageGenerationConfig, ImageGenerationResponse, ImageInput, ImageSize, LogConfig, LogEntry, LogHandler, MemoryEntry, Message, MessageContent, MessageContentPart, MessageRole, NPCConfig, PlayerInfo, RechargeConfig, RechargeEvents, RechargeModalOptions, SDKConfig, SDKMode, SchemaEntry, SetNicknameRequest, SetNicknameResponse, StreamChunk, StructuredGenerationConfig, StructuredOutputConfig, StructuredResult, TTSConfig, TTSOptions, TTSResult, TextContentPart, TokenRefreshResult, TokenScope, TokenStorageOptions, TokenValidatorOptions, TokenVerificationResult, TranscriptionConfig, TranscriptionOptions, TranscriptionResult, TranscriptionSegment, ValidatedPlayerInfo };