@ooneex/ai 1.1.14 → 1.2.1

package/dist/index.d.ts

@@ -2,30 +2,12 @@ import { Exception } from "@ooneex/exception";
 declare class AiException extends Exception {
 	constructor(message: string, key: string, data?: Record<string, unknown>);
 }
-import { AppEnv } from "@ooneex/app-env";
 import { ModelMessage } from "@tanstack/ai";
 import { LocaleType } from "@ooneex/translation";
 import { AssertType } from "@ooneex/validation";
-import { ImageGenerationResult, TTSResult } from "@tanstack/ai";
-import { anthropicText } from "@tanstack/ai-anthropic";
-import { GeminiAspectRatio, GeminiImageProviderOptions, GeminiTTSVoice, geminiText } from "@tanstack/ai-gemini";
-import { groqText } from "@tanstack/ai-groq";
-import { ollamaText } from "@tanstack/ai-ollama";
-import { OpenAITranscriptionProviderOptions, OpenAITTSFormat, OpenAITTSVoice, openaiText } from "@tanstack/ai-openai";
-import { TranscriptionResult, TranscriptionSegment, TranscriptionWord } from "@tanstack/ai";
+import { createOpenRouterText } from "@tanstack/ai-openrouter";
 type AiClassType = new (...args: any[]) => IAiChat<any>;
-type OpenAiModelType = Parameters<typeof openaiText>[0];
-type AnthropicModelType = Parameters<typeof anthropicText>[0];
-type GeminiModelType = Parameters<typeof geminiText>[0];
-type GroqModelType = Parameters<typeof groqText>[0];
-type OllamaModelType = Parameters<typeof ollamaText>[0];
-type OpenAiTTSModelType = "tts-1" | "tts-1-hd" | "gpt-4o-audio-preview";
-type OpenAiSTTModelType = "whisper-1" | "gpt-4o-transcribe" | "gpt-4o-mini-transcribe" | "gpt-4o-transcribe-diarize";
-type GeminiTTSModelType = "gemini-2.5-flash-preview-tts" | "gemini-2.5-pro-preview-tts";
-type OpenAiImageModelType = "gpt-image-1" | "gpt-image-1-mini" | "dall-e-3" | "dall-e-2";
-type OpenAiImageSizeType = "1024x1024" | "1536x1024" | "1024x1536" | "1792x1024" | "1024x1792" | "256x256" | "512x512" | "auto";
-type GeminiImageModelType = "gemini-2.5-flash-image" | "gemini-2.0-flash-preview-image-generation" | "imagen-3.0-generate-002" | "imagen-4.0-generate-001" | "imagen-4.0-fast-generate-001" | "imagen-4.0-ultra-generate-001";
-type GeminiImageSizeType = "1024x1024" | "512x512" | "1024x768" | "1536x1024" | "1792x1024" | "1920x1080" | "768x1024" | "1024x1536" | "1024x1792" | "1080x1920";
+type OpenRouterModelType = Parameters<typeof createOpenRouterText>[0];
 type GenerateQuestionOptionsType = {
 	choiceCount?: number;
 	correctChoiceCount?: number;
@@ -77,13 +59,38 @@ type AiImageSourceType = {
 	type: "url" | "data";
 	value: string;
 };
+type AiAudioSourceType = {
+	type: "url" | "data";
+	value: string;
+};
+type AiVideoResultType = {
+	jobId: string;
+	url?: string | undefined;
+	status: "pending" | "processing" | "completed" | "failed";
+	error?: string | undefined;
+};
+type AiSpeechFormatType = "mp3" | "opus" | "aac" | "flac" | "wav" | "pcm";
+type AiSpeechResultType = {
+	audio: string;
+	format: AiSpeechFormatType;
+	duration?: number | undefined;
+	contentType?: string | undefined;
+};
+type AiImageResultType = {
+	url: string;
+	revisedPrompt?: string | undefined;
+};
+type AiVideoSourceType = {
+	type: "url" | "data";
+	value: string;
+};
 type AiMessageType = {
 	role: "user" | "assistant" | "system" | "tool";
 	content: string;
 };
 type AiConfigType = {
 	apiKey?: string;
-	model?: OpenAiModelType | AnthropicModelType | GeminiModelType | GroqModelType | OllamaModelType;
+	model?: OpenRouterModelType;
 	wordCount?: number;
 	stream?: boolean;
 	language?: LocaleType;
@@ -94,80 +101,7 @@ type AiConfigType = {
 	count?: number;
 	output?: AssertType;
 };
-type OpenAiConfigType = Omit<AiConfigType, "model"> & {
-	model?: OpenAiModelType;
-};
-type AnthropicConfigType = Omit<AiConfigType, "model"> & {
-	model?: AnthropicModelType;
-};
-type GeminiConfigType = Omit<AiConfigType, "model"> & {
-	model?: GeminiModelType;
-};
-type GroqConfigType = Omit<AiConfigType, "model"> & {
-	model?: GroqModelType;
-};
-type OllamaConfigType = Omit<AiConfigType, "model" | "apiKey"> & {
-	host?: string;
-	model?: OllamaModelType;
-};
-type OpenAiTextToSpeechOptionsType = {
-	apiKey?: string;
-	model?: OpenAiTTSModelType;
-	voice?: OpenAITTSVoice;
-	format?: OpenAITTSFormat;
-	speed?: number;
-	language?: string;
-	instructions?: string;
-};
-type OpenAiSpeechToTextOptionsType = {
-	apiKey?: string;
-	model?: OpenAiSTTModelType;
-	language?: string;
-	prompt?: string;
-	responseFormat?: "json" | "text" | "srt" | "verbose_json" | "vtt";
-	modelOptions?: OpenAITranscriptionProviderOptions;
-};
-type GroqTTSModelType = "canopylabs/orpheus-v1-english" | "canopylabs/orpheus-arabic-saudi";
-type GroqTTSVoiceType = "autumn" | "diana" | "hannah" | "austin" | "daniel" | "troy" | "fahad" | "sultan" | "lulwa" | "noura";
-type GroqTTSFormatType = "wav" | "mp3" | "flac" | "ogg" | "mulaw";
-type GroqTextToSpeechOptionsType = {
-	apiKey?: string;
-	model?: GroqTTSModelType;
-	voice?: GroqTTSVoiceType;
-	format?: GroqTTSFormatType;
-	sampleRate?: number;
-};
-type GeminiTextToSpeechOptionsType = {
-	apiKey?: string;
-	model?: GeminiTTSModelType;
-	voice?: GeminiTTSVoice;
-	format?: "mp3" | "opus" | "aac" | "flac" | "wav" | "pcm";
-	speed?: number;
-	instructions?: string;
-	language?: string;
-};
-type OpenAiGenerateImageOptionsType = {
-	apiKey?: string;
-	model?: OpenAiImageModelType;
-	numberOfImages?: number;
-	size?: OpenAiImageSizeType;
-	quality?: "high" | "medium" | "low" | "auto" | "hd" | "standard";
-	background?: "transparent" | "opaque" | "auto";
-	outputFormat?: "png" | "jpeg" | "webp";
-	moderation?: "low" | "auto";
-	style?: "vivid" | "natural";
-};
-type GeminiGenerateImageOptionsType = {
-	apiKey?: string;
-	model?: GeminiImageModelType;
-	numberOfImages?: number;
-	size?: GeminiImageSizeType;
-	aspectRatio?: GeminiAspectRatio;
-	personGeneration?: "DONT_ALLOW" | "ALLOW_ADULT" | "ALLOW_ALL";
-	negativePrompt?: string;
-	addWatermark?: boolean;
-	outputMimeType?: "image/png" | "image/jpeg" | "image/webp";
-};
+type OpenRouterConfigType = AiConfigType;
 interface IAiChat<TConfig extends AiConfigType = AiConfigType> {
 	makeShorter?: (content: string, config?: TConfig) => Promise<string>;
 	makeLonger?: (content: string, config?: TConfig) => Promise<string>;
@@ -190,16 +124,35 @@ interface IAiChat<TConfig extends AiConfigType = AiConfigType> {
 	generateCaseQuestion?: (subject: string, options?: GenerateCaseQuestionOptionsType, config?: TConfig) => Promise<GenerateCaseQuestionResultType>;
 	generateFlashcard?: (subject: string, options?: GenerateFlashcardOptionsType, config?: TConfig) => Promise<GenerateFlashcardResultType>;
 	generateQuestion?: (subject: string, options?: GenerateQuestionOptionsType, config?: TConfig) => Promise<GenerateQuestionResultType>;
+	describeImage?: (source: AiImageSourceType, config?: Omit<TConfig, "output">) => Promise<string>;
 	imageToMarkdown?: (source: AiImageSourceType, config?: Omit<TConfig, "output">) => Promise<string>;
+	imageToText?: (source: AiImageSourceType, config?: Omit<TConfig, "output">) => Promise<string>;
+	speechToText?: (source: AiAudioSourceType, config?: Omit<TConfig, "output">) => Promise<string>;
+	videoToText?: (source: AiVideoSourceType, config?: Omit<TConfig, "output">) => Promise<string>;
+	imageToImage?: (source: AiImageSourceType, prompt: string, config?: Omit<TConfig, "output"> & {
+		size?: string;
+		quality?: "standard" | "hd";
+	}) => Promise<AiImageResultType>;
+	textToImage?: (prompt: string, config?: Omit<TConfig, "output"> & {
+		size?: string;
+		quality?: "standard" | "hd";
+	}) => Promise<AiImageResultType>;
+	textToSpeech?: (text: string, config?: Omit<TConfig, "output"> & {
+		voice?: string;
+		format?: AiSpeechFormatType;
+		speed?: number;
+	}) => Promise<AiSpeechResultType>;
+	textToVideo?: (prompt: string, config?: Omit<TConfig, "output">) => Promise<AiVideoResultType>;
+	getVideoStatus?: (jobId: string, config?: Omit<TConfig, "output">) => Promise<AiVideoResultType>;
 	run: <T>(content: string, config?: TConfig) => Promise<T>;
 	runStream: (content: string, config?: TConfig) => AsyncGenerator<string, void, unknown>;
 }
 declare abstract class BaseAi<TConfig extends AiConfigType> implements IAiChat<TConfig> {
-	protected abstract createChatAdapter(config?: TConfig): any;
-	protected abstract createRunAdapter(config?: TConfig): any;
+	protected abstract createChatAdapter(config?: TConfig, task?: string): any;
+	protected abstract createRunAdapter(config?: TConfig, task?: string): any;
 	protected buildPrompt(instruction: string, config?: TConfig): string;
 	protected toMessages(messages: AiMessageType[]): ModelMessage[];
-	protected executeChat(content: string, systemPrompt: string, config?: TConfig): Promise<string>;
+	protected executeChat(content: string, systemPrompt: string, config?: TConfig, task?: string): Promise<string>;
 	makeShorter(content: string, config?: Omit<TConfig, "output">): Promise<string>;
 	makeLonger(content: string, config?: Omit<TConfig, "output">): Promise<string>;
 	summarize(content: string, config?: Omit<TConfig, "output">): Promise<string>;
@@ -221,60 +174,53 @@ declare abstract class BaseAi<TConfig extends AiConfigType> implements IAiChat<T
 	generateCaseQuestion(subject: string, options?: GenerateCaseQuestionOptionsType, config?: TConfig): Promise<GenerateCaseQuestionResultType>;
 	generateFlashcard(subject: string, options?: GenerateFlashcardOptionsType, config?: TConfig): Promise<GenerateFlashcardResultType>;
 	generateQuestion(subject: string, options?: GenerateQuestionOptionsType, config?: TConfig): Promise<GenerateQuestionResultType>;
+	describeImage(source: AiImageSourceType, config?: Omit<TConfig, "output">): Promise<string>;
 	imageToMarkdown(source: AiImageSourceType, config?: Omit<TConfig, "output">): Promise<string>;
+	imageToText(source: AiImageSourceType, config?: Omit<TConfig, "output">): Promise<string>;
+	speechToText(source: AiAudioSourceType, config?: Omit<TConfig, "output">): Promise<string>;
+	videoToText(source: AiVideoSourceType, config?: Omit<TConfig, "output">): Promise<string>;
+	abstract textToImage(prompt: string, config?: Omit<TConfig, "output"> & {
+		size?: string;
+		quality?: "standard" | "hd";
+	}): Promise<AiImageResultType>;
+	abstract textToSpeech(text: string, config?: Omit<TConfig, "output"> & {
+		voice?: string;
+		format?: AiSpeechFormatType;
+		speed?: number;
+	}): Promise<AiSpeechResultType>;
+	abstract textToVideo(prompt: string, config?: Omit<TConfig, "output">): Promise<AiVideoResultType>;
+	abstract getVideoStatus(jobId: string, config?: Omit<TConfig, "output">): Promise<AiVideoResultType>;
 	run<T>(prompt: string, config?: Omit<TConfig, "prompt">): Promise<T>;
 	runStream(prompt: string, config?: Omit<TConfig, "prompt" | "output">): AsyncGenerator<string, void, unknown>;
 }
-declare class AnthropicAi extends BaseAi<AnthropicConfigType> {
-	private readonly env;
-	constructor(env: AppEnv);
-	private getApiKey;
-	protected createChatAdapter(config?: AnthropicConfigType);
-	protected createRunAdapter(config?: AnthropicConfigType);
-}
 import { EContainerScope } from "@ooneex/container";
 declare const decorator: {
 	ai: (scope?: EContainerScope) => (target: AiClassType) => void;
 };
-import { AppEnv as AppEnv2 } from "@ooneex/app-env";
-import { ImageGenerationResult as ImageGenerationResult2, TTSResult as TTSResult2 } from "@tanstack/ai";
-declare class GeminiAi extends BaseAi<GeminiConfigType> {
-	private readonly env;
-	constructor(env: AppEnv2);
-	private getApiKey;
-	protected createChatAdapter(config?: GeminiConfigType);
-	protected createRunAdapter(config?: GeminiConfigType);
-	textToSpeech(text: string, options?: GeminiTextToSpeechOptionsType): Promise<TTSResult2>;
-	generateImage(prompt: string, options?: GeminiGenerateImageOptionsType): Promise<ImageGenerationResult2>;
-}
-import { AppEnv as AppEnv3 } from "@ooneex/app-env";
-import { TTSResult as TTSResult3 } from "@tanstack/ai";
-declare class GroqAi extends BaseAi<GroqConfigType> {
-	private readonly env;
-	constructor(env: AppEnv3);
-	private getApiKey;
-	protected createChatAdapter(config?: GroqConfigType);
-	protected createRunAdapter(config?: GroqConfigType);
-	textToSpeech(text: string, options?: GroqTextToSpeechOptionsType): Promise<TTSResult3>;
-}
-import { AppEnv as AppEnv4 } from "@ooneex/app-env";
-declare class OllamaAi extends BaseAi<OllamaConfigType> {
-	private readonly env;
-	constructor(env: AppEnv4);
-	private getHost;
-	protected createChatAdapter(config?: OllamaConfigType);
-	protected createRunAdapter(config?: OllamaConfigType);
-}
-import { AppEnv as AppEnv5 } from "@ooneex/app-env";
-import { ImageGenerationResult as ImageGenerationResult3, TranscriptionResult as TranscriptionResult2, TTSResult as TTSResult4 } from "@tanstack/ai";
-declare class OpenAi extends BaseAi<OpenAiConfigType> {
+import { AppEnv } from "@ooneex/app-env";
+declare class OpenRouterAi extends BaseAi<OpenRouterConfigType> {
 	private readonly env;
-	constructor(env: AppEnv5);
+	private static readonly DEFAULT_MODELS;
+	constructor(env: AppEnv);
 	private getApiKey;
-	protected createChatAdapter(config?: OpenAiConfigType);
-	protected createRunAdapter(config?: OpenAiConfigType);
-	textToSpeech(text: string, options?: OpenAiTextToSpeechOptionsType): Promise<TTSResult4>;
-	speechToText(audio: string | File | Blob | ArrayBuffer, options?: OpenAiSpeechToTextOptionsType): Promise<TranscriptionResult2>;
-	generateImage(prompt: string, options?: OpenAiGenerateImageOptionsType): Promise<ImageGenerationResult3>;
+	private fetchWithRetry;
+	private getDefaultModel;
+	textToImage(prompt: string, config?: Omit<OpenRouterConfigType, "output"> & {
+		size?: string;
+		quality?: "standard" | "hd";
+	}): Promise<AiImageResultType>;
+	imageToImage(source: AiImageSourceType, prompt: string, config?: Omit<OpenRouterConfigType, "output"> & {
+		size?: string;
+		quality?: "standard" | "hd";
+	}): Promise<AiImageResultType>;
+	textToSpeech(text: string, config?: Omit<OpenRouterConfigType, "output"> & {
+		voice?: string;
+		format?: AiSpeechFormatType;
+		speed?: number;
+	}): Promise<AiSpeechResultType>;
+	textToVideo(prompt: string, config?: Omit<OpenRouterConfigType, "output">): Promise<AiVideoResultType>;
+	getVideoStatus(jobId: string, config?: Omit<OpenRouterConfigType, "output">): Promise<AiVideoResultType>;
+	protected createChatAdapter(config?: OpenRouterConfigType, task?: string);
+	protected createRunAdapter(config?: OpenRouterConfigType, task?: string);
 }
-export { decorator, TranscriptionWord, TranscriptionSegment, TranscriptionResult, TTSResult, OpenAiTextToSpeechOptionsType, OpenAiTTSModelType, OpenAiSpeechToTextOptionsType, OpenAiSTTModelType, OpenAiModelType, OpenAiImageSizeType, OpenAiImageModelType, OpenAiGenerateImageOptionsType, OpenAiConfigType, OpenAi, OpenAITranscriptionProviderOptions, OpenAITTSVoice, OpenAITTSFormat, OllamaModelType, OllamaConfigType, OllamaAi, ImageGenerationResult, IAiChat, GroqTextToSpeechOptionsType, GroqTTSVoiceType, GroqTTSModelType, GroqTTSFormatType, GroqModelType, GroqConfigType, GroqAi, GenerateQuestionResultType, GenerateQuestionOptionsType, GenerateQuestionChoiceType, GenerateFlashcardResultType, GenerateFlashcardOptionsType, GenerateCaseQuestionResultType, GenerateCaseQuestionOptionsType, GeminiTextToSpeechOptionsType, GeminiTTSVoice, GeminiTTSModelType, GeminiModelType, GeminiImageSizeType, GeminiImageProviderOptions, GeminiImageModelType, GeminiGenerateImageOptionsType, GeminiConfigType, GeminiAspectRatio, GeminiAi, CaseQuestionType, CaseQuestionChoiceType, BaseAi, AnthropicModelType, AnthropicConfigType, AnthropicAi, AiToneType, AiMessageType, AiImageSourceType, AiException, AiConfigType, AiClassType };
+export { decorator, OpenRouterModelType, OpenRouterConfigType, OpenRouterAi, IAiChat, GenerateQuestionResultType, GenerateQuestionOptionsType, GenerateQuestionChoiceType, GenerateFlashcardResultType, GenerateFlashcardOptionsType, GenerateCaseQuestionResultType, GenerateCaseQuestionOptionsType, CaseQuestionType, CaseQuestionChoiceType, BaseAi, AiVideoSourceType, AiVideoResultType, AiToneType, AiSpeechResultType, AiSpeechFormatType, AiMessageType, AiImageSourceType, AiImageResultType, AiException, AiConfigType, AiClassType, AiAudioSourceType };