unspeech 0.1.14 → 0.1.20

package/README.md

@@ -51,6 +51,7 @@ import {
   createUnAlibabaCloud,
   createUnElevenLabs,
   createUnMicrosoft,
+  createUnMinimax,
   createUnSpeech,
   createUnVolcengine,
 } from 'unspeech'
@@ -62,6 +63,7 @@ When using
 - [Alibaba Cloud Model Studio / 阿里云百炼 / CosyVoice](https://www.alibabacloud.com/en/product/modelstudio)
 - [Volcano Engine / 火山引擎语音技术](https://www.volcengine.com/product/voice-tech)
 - [ElevenLabs](https://elevenlabs.io/docs/api-reference/text-to-speech/convert)
+- [MiniMax](https://platform.minimaxi.com/docs/guides/speech-t2a-http)
 
 providers, [SSML](https://learn.microsoft.com/en-us/azure/ai-services/speech-service/speech-synthesis-markup) is supported to control in fine grain level for pitch, volume, rate, etc.
 

package/dist/index.d.mts

@@ -353,6 +353,210 @@ interface UnElevenLabsOptions {
  */
 declare function createUnElevenLabs(apiKey: string, baseURL?: string): SpeechProviderWithExtraOptions<"eleven_english_sts_v2" | "eleven_flash_v2" | "eleven_flash_v2_5" | "eleven_multilingual_sts_v2" | "eleven_multilingual_v2", UnElevenLabsOptions> & VoiceProviderWithExtraOptions<UnElevenLabsOptions>;
 //#endregion
+//#region src/backend/fishaudio.d.ts
+/** @see {@link https://docs.fish.audio/api-reference/endpoint/openapi-v1/text-to-speech} */
+interface UnFishAudioOptions {
+  /**
+   * Text segment length used when splitting long input.
+   * Range: 100 to 300.
+   *
+   * @default 200
+   */
+  chunkLength?: number;
+  /**
+   * Latency mode. `balanced` reduces latency at a slight cost to stability.
+   *
+   * @default 'normal'
+   */
+  latency?: 'balanced' | 'normal';
+  /**
+   * MP3 bitrate in kbps, only used with the mp3 format.
+   *
+   * @default 128
+   */
+  mp3Bitrate?: 64 | 128 | 192;
+  /**
+   * Normalize text (e.g. spell out numbers) for better stability.
+   *
+   * @default true
+   */
+  normalize?: boolean;
+  /**
+   * Speech pacing controls.
+   */
+  prosody?: {
+    /**
+     * Speed multiplier, 1.0 is normal speed.
+     *
+     * @default 1
+     */
+    speed?: number;
+    /**
+     * Volume gain in dB, 0 is unchanged.
+     *
+     * @default 0
+     */
+    volume?: number;
+  };
+  /**
+   * Output sample rate in Hz.
+   */
+  sampleRate?: number;
+  /**
+   * Sampling temperature; lower is more deterministic.
+   *
+   * @default 0.7
+   */
+  temperature?: number;
+  /**
+   * Nucleus sampling threshold.
+   *
+   * @default 0.7
+   */
+  topP?: number;
+}
+/**
+ * [Fish Audio](https://fish.audio/) provider for
+ * [UnSpeech](https://github.com/moeru-ai/unspeech).
+ *
+ * The OpenAI `voice` parameter maps to a Fish Audio reference voice model id
+ * (the id in a fish.audio voice page URL).
+ *
+ * @param apiKey - Fish Audio API key
+ * @param baseURL - UnSpeech Instance URL
+ * @returns SpeechProviderWithExtraOptions
+ */
+declare function createUnFishAudio(apiKey: string, baseURL?: string): SpeechProviderWithExtraOptions<`fishaudio/${string}`, UnFishAudioOptions> & VoiceProviderWithExtraOptions<UnFishAudioOptions>;
+//#endregion
+//#region src/backend/minimax.d.ts
+/**
+ * MiniMax TTS API options
+ * @see https://platform.minimaxi.com/docs/guides/speech-t2a-http
+ */
+interface UnMinimaxOptions {
+  /**
+   * Speech speed. Range: 0.5-2.0
+   * @default 1.0
+   */
+  speed?: number;
+  /**
+   * Volume. Range: 0-10
+   * @default 1.0
+   */
+  vol?: number;
+  /**
+   * Pitch adjustment. Range: -12 to 12
+   * @default 0
+   */
+  pitch?: number;
+  /**
+   * Emotion setting
+   * @see https://platform.minimaxi.com/docs/guides/speech-t2a-http#emotion
+   * @example "happy" | "sad" | "angry" | "fearful" | "disgusted" | "surprised" | "calm" | "fluent" | "whisper"
+   */
+  emotion?: string;
+  /**
+   * Enable streaming output
+   * @default false
+   */
+  stream?: boolean;
+  /**
+   * Sample rate for audio output
+   * @example 8000 | 16000 | 22050 | 24000 | 32000 | 44100
+   */
+  sampleRate?: number;
+  /**
+   * Audio bitrate
+   * @example 32000 | 64000 | 128000 | 256000
+   */
+  bitrate?: number;
+  /**
+   * Audio format
+   * @example "mp3" | "pcm" | "flac" | "wav"
+   * @default "mp3"
+   */
+  format?: 'mp3' | 'pcm' | 'flac' | 'wav';
+  /**
+   * Audio channel
+   * @example 1 | 2
+   * @default 1
+   */
+  channel?: number;
+}
+/**
+ * MiniMax TTS models
+ * @see https://platform.minimaxi.com/docs/guides/speech-t2a-http#model
+ */
+type MinimaxModel = 'speech-2.8-hd' | 'speech-2.8-turbo' | 'speech-2.6-hd' | 'speech-2.6-turbo' | 'speech-02-hd' | 'speech-02-turbo' | 'speech-01-hd' | 'speech-01-turbo';
+/**
+ * [MiniMax](https://platform.minimaxi.com/) provider for [UnSpeech](https://github.com/moeru-ai/unspeech)
+ *
+ * @param apiKey - MiniMax API Key
+ * @param baseURL - UnSpeech Instance URL
+ * @returns SpeechProviderWithExtraOptions & VoiceProviderWithExtraOptions
+ */
+declare function createUnMinimax(apiKey: string, baseURL?: string): SpeechProviderWithExtraOptions<"minimax/speech-2.8-hd" | "minimax/speech-2.8-turbo" | "minimax/speech-2.6-hd" | "minimax/speech-2.6-turbo" | "minimax/speech-02-hd" | "minimax/speech-02-turbo" | "minimax/speech-01-hd" | "minimax/speech-01-turbo", UnMinimaxOptions> & VoiceProviderWithExtraOptions<UnMinimaxOptions>;
+//#endregion
+//#region src/backend/stepfun.d.ts
+interface UnStepfunOptions {
+  /**
+   * Selects a StepFun-defined API endpoint without exposing an arbitrary URL.
+   *
+   * Use `step-plan` for requests covered by a Step Plan subscription. Omit
+   * this option, or use `default`, for StepFun's default speech endpoint.
+   *
+   * @default 'default'
+   */
+  endpointProfile?: 'default' | 'step-plan';
+  /**
+   * The language that StepFun reads the text in, for example `en` or `ja`.
+   *
+   * Without it, a short or mixed text, such as a digit next to Latin letters,
+   * can be read in the language of the voice.
+   */
+  language?: 'zh' | 'en' | 'ja' | 'ko' | 'fr' | 'es' | string;
+  /**
+   * Audio volume. Range: 0.1 to 2.0.
+   *
+   * @default 1
+   */
+  volume?: number;
+  /**
+   * Voice labels for `step-tts-2` and `step-tts-mini`.
+   *
+   * Not supported by `stepaudio-2.5-tts`; use `instruction` instead.
+   */
+  voiceLabel?: {
+    language?: '粤语' | '四川话' | '日语' | string;
+    emotion?: string;
+    style?: string;
+  };
+  /**
+   * Global natural-language guidance for `stepaudio-2.5-tts`.
+   */
+  instruction?: string;
+  /**
+   * Output sample rate.
+   *
+   * @default 24000
+   */
+  sampleRate?: 8000 | 16000 | 22050 | 24000 | 48000 | number;
+  pronunciationMap?: {
+    tone: string[];
+  };
+  streamFormat?: 'audio' | 'sse';
+  markdownFilter?: boolean;
+}
+/**
+ * [StepFun / 阶跃星辰](https://platform.stepfun.com/) provider for
+ * [UnSpeech](https://github.com/moeru-ai/unspeech).
+ *
+ * @param apiKey - StepFun API key
+ * @param baseURL - UnSpeech Instance URL
+ * @returns SpeechProviderWithExtraOptions
+ */
+declare function createUnStepfun(apiKey: string, baseURL?: string): SpeechProviderWithExtraOptions<`stepfun/${string}`, UnStepfunOptions> & VoiceProviderWithExtraOptions<UnStepfunOptions>;
+//#endregion
 //#region src/backend/volcengine.d.ts
 interface UnVolcengineOptions {
   app?: {
@@ -456,7 +660,7 @@ declare function createUnVolcengine(apiKey: string, baseURL?: string): SpeechPro
 //#endregion
 //#region src/backend/index.d.ts
 /** @see {@link https://github.com/moeru-ai/unspeech} */
-declare function createUnSpeech(apiKey: string, baseURL?: string): _xsai_ext_providers_utils0.SpeechProviderWithExtraOptions<`alibaba/${string}` | `aliyun/${string}` | `deepgram/${string}` | `elevenlabs/${string}` | `koemotion/${string}` | `openai/${string}` | `volcano/${string}` | `volcengine/${string}`, UnSpeechOptions> & VoiceProviderWithExtraOptions<{
+declare function createUnSpeech(apiKey: string, baseURL?: string): _xsai_ext_providers_utils0.SpeechProviderWithExtraOptions<`alibaba/${string}` | `aliyun/${string}` | `deepgram/${string}` | `elevenlabs/${string}` | `fishaudio/${string}` | `koemotion/${string}` | `minimax/${string}` | `openai/${string}` | `stepfun/${string}` | `volcano/${string}` | `volcengine/${string}`, UnSpeechOptions> & VoiceProviderWithExtraOptions<{
   appId: string;
   backend: "volcano";
 } | {
@@ -466,7 +670,7 @@ declare function createUnSpeech(apiKey: string, baseURL?: string): _xsai_ext_pro
   backend: "azure" | "microsoft";
   region: MicrosoftRegions | string;
 } | {
-  backend: "ali" | "alibaba" | "alibaba-model-studio" | "aliyun" | "bailian" | "deepgram" | "elevenlabs" | "koemotion" | "openai";
+  backend: "ali" | "alibaba" | "alibaba-model-studio" | "aliyun" | "bailian" | "deepgram" | "elevenlabs" | "fish" | "fish-audio" | "fishaudio" | "koemotion" | "minimax" | "openai" | "step" | "stepfun";
 }>;
 //#endregion
 //#region src/utils/generate-speech-response.d.ts
@@ -544,4 +748,4 @@ interface ListVoicesResponse {
 }
 declare function listVoices(options: ListVoicesOptions): Promise<Voice[]>;
 //#endregion
-export { DEFAULT_MICROSOFT_OUTPUT_FORMAT, GenerateSpeechResponseOptions, GenerateSpeechResponseResult, ListVoicesOptions, ListVoicesResponse, MicrosoftRegions, UnAlibabaCloudOptions, UnDeepgramOptions, UnElevenLabsOptions, UnMicrosoftOptionAutoSSML, UnMicrosoftOptionCommon, UnMicrosoftOptionCustomSSML, UnMicrosoftOptions, UnSpeechAPIError, UnSpeechAPIErrorOptions, UnSpeechOptions, UnVolcengineOptions, Voice, VoiceFormat, VoiceLanguage, VoiceProvider, VoiceProviderWithExtraOptions, buildMicrosoftSsml, createUnAlibabaCloud, createUnDeepgram, createUnElevenLabs, createUnMicrosoft, createUnSpeech, createUnVolcengine, escapeMicrosoftSsmlText, generateSpeechResponse, inferMicrosoftContentType, isMicrosoftVoiceId, listVoices, microsoftSpeedToProsodyRate, resolveMicrosoftOutputFormat };
+export { DEFAULT_MICROSOFT_OUTPUT_FORMAT, GenerateSpeechResponseOptions, GenerateSpeechResponseResult, ListVoicesOptions, ListVoicesResponse, MicrosoftRegions, MinimaxModel, UnAlibabaCloudOptions, UnDeepgramOptions, UnElevenLabsOptions, UnFishAudioOptions, UnMicrosoftOptionAutoSSML, UnMicrosoftOptionCommon, UnMicrosoftOptionCustomSSML, UnMicrosoftOptions, UnMinimaxOptions, UnSpeechAPIError, UnSpeechAPIErrorOptions, UnSpeechOptions, UnStepfunOptions, UnVolcengineOptions, Voice, VoiceFormat, VoiceLanguage, VoiceProvider, VoiceProviderWithExtraOptions, buildMicrosoftSsml, createUnAlibabaCloud, createUnDeepgram, createUnElevenLabs, createUnFishAudio, createUnMicrosoft, createUnMinimax, createUnSpeech, createUnStepfun, createUnVolcengine, escapeMicrosoftSsmlText, generateSpeechResponse, inferMicrosoftContentType, isMicrosoftVoiceId, listVoices, microsoftSpeedToProsodyRate, resolveMicrosoftOutputFormat };

package/dist/index.mjs

@@ -279,6 +279,45 @@ function createUnMicrosoft(apiKey, baseURL = "http://localhost:5933/v1/") {
 	} });
 }
 
+//#endregion
+//#region src/backend/minimax.ts
+/**
+* [MiniMax](https://platform.minimaxi.com/) provider for [UnSpeech](https://github.com/moeru-ai/unspeech)
+*
+* @param apiKey - MiniMax API Key
+* @param baseURL - UnSpeech Instance URL
+* @returns SpeechProviderWithExtraOptions & VoiceProviderWithExtraOptions
+*/
+function createUnMinimax(apiKey, baseURL = "http://localhost:5933/v1/") {
+	const toUnSpeechOptions = ({ speed, vol, pitch, emotion, stream, sampleRate, bitrate, format, channel }) => ({ extraBody: objCamelToSnake({
+		speed,
+		vol,
+		pitch,
+		emotion,
+		stream,
+		sampleRate,
+		bitrate,
+		format,
+		channel
+	}) });
+	return merge({ speech: (model, options) => ({
+		...options ? toUnSpeechOptions(options) : {},
+		apiKey,
+		baseURL,
+		model
+	}) }, { voice: (options) => {
+		let adjustedBaseURL = baseURL;
+		if (adjustedBaseURL.endsWith("v1/")) adjustedBaseURL = adjustedBaseURL.slice(0, -3);
+		else if (adjustedBaseURL.endsWith("v1")) adjustedBaseURL = adjustedBaseURL.slice(0, -2);
+		return {
+			query: "provider=minimax",
+			...options ? toUnSpeechOptions(options) : {},
+			apiKey,
+			baseURL: adjustedBaseURL
+		};
+	} });
+}
+
 //#endregion
 //#region src/backend/volcengine.ts
 /**
@@ -448,4 +487,4 @@ async function listVoices(options) {
 }
 
 //#endregion
-export { UnSpeechAPIError, buildMicrosoftSsml, createUnAlibabaCloud, createUnDeepgram, createUnElevenLabs, createUnMicrosoft, createUnSpeech, createUnVolcengine, escapeMicrosoftSsmlText, generateSpeechResponse, inferMicrosoftContentType, isMicrosoftVoiceId, listVoices, microsoftSpeedToProsodyRate, resolveMicrosoftOutputFormat };
+export { UnSpeechAPIError, buildMicrosoftSsml, createUnAlibabaCloud, createUnDeepgram, createUnElevenLabs, createUnMicrosoft, createUnMinimax, createUnSpeech, createUnVolcengine, escapeMicrosoftSsmlText, generateSpeechResponse, inferMicrosoftContentType, isMicrosoftVoiceId, listVoices, microsoftSpeedToProsodyRate, resolveMicrosoftOutputFormat };

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "unspeech",
   "type": "module",
-  "version": "0.1.14",
+  "version": "0.1.20",
   "description": "Client SDK for unSpeech",
   "author": "Moeru AI",
   "license": "MIT",