feat: support native tts engine on Android (#1387)

2025-06-12 23:56:14 +08:00
parent 28c6c1ec58
commit 658969e97a
52 changed files with 931 additions and 755 deletions
@@ -236,7 +236,7 @@
  "Border Color": "لون الحدود",
  "Border Frame": "إطار الحدود",
  "Show Header": "إظهار الترويسة",
-  "Show Footer": "إظهار التذييل"
+  "Show Footer": "إظهار التذييل",
  "Small": "صغير",
  "Large": "كبير",
  "Auto": "تلقائي",
@@ -357,5 +357,11 @@
  "Quota Exceeded": "تجاوزت الحصة المخصصة",
  "{{percentage}}% of Daily Translation Characters Used.": "تم استخدام ما نسبته {{percentage}}% من إجمالي أحرف الترجمة المسموح بها يوميًا.",
  "Translation Characters": "أحرف الترجمة",
-  "Daily translation quota reached. Select another translate service to proceed.": "تم استنفاد حصة الترجمة اليومية. قم بتحديد خدمة ترجمة أخرى للاستمرار."
+  "Daily translation quota reached. Select another translate service to proceed.": "تم استنفاد حصة الترجمة اليومية. قم بتحديد خدمة ترجمة أخرى للاستمرار.",
+  "{{engine}}: {{count}} voices_zero": "{{engine}}: لا توجد أصوات",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: صوت واحد",
+  "{{engine}}: {{count}} voices_two": "{{engine}}: صوتان",
+  "{{engine}}: {{count}} voices_few": "{{engine}}: {{count}} أصوات",
+  "{{engine}}: {{count}} voices_many": "{{engine}}: {{count}} صوتاً",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} صوت"
 }
@@ -349,5 +349,7 @@
  "Quota Exceeded": "Limit überschritten",
  "{{percentage}}% of Daily Translation Characters Used.": "{{percentage}}% der täglichen Übersetzungszeichen verwendet.",
  "Translation Characters": "Übersetzungszeichen",
-  "Daily translation quota reached. Select another translate service to proceed.": "Das tägliche Übersetzungslimit wurde erreicht. Wählen Sie einen anderen Übersetzungsdienst, um fortzufahren."
+  "Daily translation quota reached. Select another translate service to proceed.": "Das tägliche Übersetzungslimit wurde erreicht. Wählen Sie einen anderen Übersetzungsdienst, um fortzufahren.",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: {{count}} Stimme",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} Stimmen"
 }
@@ -349,5 +349,7 @@
  "Quota Exceeded": "Υπέρβαση ορίου",
  "{{percentage}}% of Daily Translation Characters Used.": "{{percentage}}% των ημερήσιων χαρακτήρων μετάφρασης χρησιμοποιήθηκαν.",
  "Translation Characters": "Χαρακτήρες μετάφρασης",
-  "Daily translation quota reached. Select another translate service to proceed.": "Έχετε φτάσει το ημερήσιο όριο χαρακτήρων μετάφρασης. Επιλέξτε άλλη υπηρεσία μετάφρασης για να συνεχίσετε."
+  "Daily translation quota reached. Select another translate service to proceed.": "Έχετε φτάσει το ημερήσιο όριο χαρακτήρων μετάφρασης. Επιλέξτε άλλη υπηρεσία μετάφρασης για να συνεχίσετε.",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: {{count}} φωνή",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} φωνές"
 }
@@ -351,5 +351,8 @@
  "Quota Exceeded": "Límite superado",
  "{{percentage}}% of Daily Translation Characters Used.": "{{percentage}}% de caracteres de traducción diarios utilizados.",
  "Translation Characters": "Caracteres de traducción",
-  "Daily translation quota reached. Select another translate service to proceed.": "Cuota diaria de traducción alcanzada. Selecciona otro servicio de traducción para continuar."
+  "Daily translation quota reached. Select another translate service to proceed.": "Cuota diaria de traducción alcanzada. Selecciona otro servicio de traducción para continuar.",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: {{count}} voz",
+  "{{engine}}: {{count}} voices_many": "{{engine}}: {{count}} voces",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} voces"
 }
@@ -351,5 +351,8 @@
  "Quota Exceeded": "Quota dépassé",
  "{{percentage}}% of Daily Translation Characters Used.": "{{percentage}}% des caractères de traduction quotidiens utilisés.",
  "Translation Characters": "Caractères de traduction",
-  "Daily translation quota reached. Select another translate service to proceed.": "Quota de traduction quotidien atteint. Sélectionnez un autre service de traduction pour continuer."
+  "Daily translation quota reached. Select another translate service to proceed.": "Quota de traduction quotidien atteint. Sélectionnez un autre service de traduction pour continuer.",
+  "{{engine}}: {{count}} voices_one": "{{engine}} : {{count}} voix",
+  "{{engine}}: {{count}} voices_many": "{{engine}} : {{count}} voix",
+  "{{engine}}: {{count}} voices_other": "{{engine}} : {{count}} voix"
 }
@@ -349,5 +349,7 @@
  "Quota Exceeded": "सीमा पार",
  "{{percentage}}% of Daily Translation Characters Used.": "{{percentage}}% दैनिक अनुवाद वर्णों का उपयोग किया गया है।",
  "Translation Characters": "अनुवाद वर्ण",
-  "Daily translation quota reached. Select another translate service to proceed.": "दैनिक अनुवाद कोटा पूरा हो गया है। आगे बढ़ने के लिए कोई अन्य अनुवाद सेवा चुनें।"
+  "Daily translation quota reached. Select another translate service to proceed.": "दैनिक अनुवाद कोटा पूरा हो गया है। आगे बढ़ने के लिए कोई अन्य अनुवाद सेवा चुनें।",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: {{count}} आवाज़",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} आवाज़ें"
 }
@@ -347,5 +347,6 @@
  "Quota Exceeded": "Kuota terlampaui",
  "{{percentage}}% of Daily Translation Characters Used.": "{{percentage}}% dari Karakter Terjemahan Harian Digunakan.",
  "Translation Characters": "Karakter Terjemahan",
-  "Daily translation quota reached. Select another translate service to proceed.": "Kuota terjemahan harian tercapai. Pilih layanan terjemahan lain untuk melanjutkan."
+  "Daily translation quota reached. Select another translate service to proceed.": "Kuota terjemahan harian tercapai. Pilih layanan terjemahan lain untuk melanjutkan.",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} suara"
 }
@@ -351,5 +351,8 @@
  "Quota Exceeded": "Limite superato",
  "{{percentage}}% of Daily Translation Characters Used.": "{{percentage}}% dei caratteri di traduzione giornalieri utilizzati.",
  "Translation Characters": "Caratteri di traduzione",
-  "Daily translation quota reached. Select another translate service to proceed.": "Quota di traduzione giornaliera raggiunta. Seleziona un altro servizio di traduzione per procedere."
+  "Daily translation quota reached. Select another translate service to proceed.": "Quota di traduzione giornaliera raggiunta. Seleziona un altro servizio di traduzione per procedere.",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: {{count}} voce",
+  "{{engine}}: {{count}} voices_many": "{{engine}}: {{count}} voci",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} voci"
 }
@@ -347,5 +347,6 @@
  "Quota Exceeded": "上限超過",
  "{{percentage}}% of Daily Translation Characters Used.": "日次翻訳文字数の使用量は{{percentage}}%です。",
  "Translation Characters": "翻訳文字数",
-  "Daily translation quota reached. Select another translate service to proceed.": "日次翻訳クォータに達しました。続行するには別の翻訳サービスを選択してください。"
+  "Daily translation quota reached. Select another translate service to proceed.": "日次翻訳クォータに達しました。続行するには別の翻訳サービスを選択してください。",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}}音声"
 }
@@ -347,5 +347,6 @@
  "Quota Exceeded": "한도 초과",
  "{{percentage}}% of Daily Translation Characters Used.": "일일 번역 문자 사용량 {{percentage}}%.",
  "Translation Characters": "번역 문자",
-  "Daily translation quota reached. Select another translate service to proceed.": "일일 번역 할당량에 도달했습니다. 계속 진행하려면 다른 번역 서비스를 선택하세요."
+  "Daily translation quota reached. Select another translate service to proceed.": "일일 번역 할당량에 도달했습니다. 계속 진행하려면 다른 번역 서비스를 선택하세요.",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}}개의 음성"
 }
@@ -349,5 +349,7 @@
  "Quota Exceeded": "Limiet overschreden",
  "{{percentage}}% of Daily Translation Characters Used.": "{{percentage}}% van de dagelijkse vertaaltekens gebruikt.",
  "Translation Characters": "Vertaaltekens",
-  "Daily translation quota reached. Select another translate service to proceed.": "Dagelijkse vertaalquotum bereikt. Selecteer een andere vertaalservice om door te gaan."
+  "Daily translation quota reached. Select another translate service to proceed.": "Dagelijkse vertaalquotum bereikt. Selecteer een andere vertaalservice om door te gaan.",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: {{count}} stem",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} stemmen"
 }
@@ -353,5 +353,9 @@
  "Quota Exceeded": "Limit przekroczony",
  "{{percentage}}% of Daily Translation Characters Used.": "{{percentage}}% użytych znaków tłumaczenia dziennego.",
  "Translation Characters": "Znaki tłumaczenia",
-  "Daily translation quota reached. Select another translate service to proceed.": "Przekroczono dzienny limit znaków tłumaczenia. Wybierz inną usługę tłumaczenia, aby kontynuować."
+  "Daily translation quota reached. Select another translate service to proceed.": "Przekroczono dzienny limit znaków tłumaczenia. Wybierz inną usługę tłumaczenia, aby kontynuować.",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: {{count}} głos",
+  "{{engine}}: {{count}} voices_few": "{{engine}}: {{count}} głosy",
+  "{{engine}}: {{count}} voices_many": "{{engine}}: {{count}} głosów",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} głosu"
 }
@@ -351,5 +351,8 @@
  "Quota Exceeded": "Limite excedido",
  "{{percentage}}% of Daily Translation Characters Used.": "{{percentage}}% dos caracteres de tradução diários usados.",
  "Translation Characters": "Caracteres de Tradução",
-  "Daily translation quota reached. Select another translate service to proceed.": "Quota diária de tradução atingida. Selecione outro serviço de tradução para continuar."
+  "Daily translation quota reached. Select another translate service to proceed.": "Quota diária de tradução atingida. Selecione outro serviço de tradução para continuar.",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: {{count}} voz",
+  "{{engine}}: {{count}} voices_many": "{{engine}}: {{count}} vozes",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} vozes"
 }
@@ -353,5 +353,9 @@
  "Quota Exceeded": "Лимит превышен",
  "{{percentage}}% of Daily Translation Characters Used.": "Использовано {{percentage}}% от суточного лимита символов перевода.",
  "Translation Characters": "Символы перевода",
-  "Daily translation quota reached. Select another translate service to proceed.": "Достигнут суточный лимит символов перевода. Выберите другой сервис перевода для продолжения."
+  "Daily translation quota reached. Select another translate service to proceed.": "Достигнут суточный лимит символов перевода. Выберите другой сервис перевода для продолжения.",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: {{count}} голос",
+  "{{engine}}: {{count}} voices_few": "{{engine}}: {{count}} голоса",
+  "{{engine}}: {{count}} voices_many": "{{engine}}: {{count}} голосов",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} голоса"
 }
@@ -349,5 +349,7 @@
  "Quota Exceeded": "Kota aşıldı",
  "{{percentage}}% of Daily Translation Characters Used.": "Çeviri için günlük karakterlerin %{{percentage}}'si kullanıldı.",
  "Translation Characters": "Çeviri Karakterleri",
-  "Daily translation quota reached. Select another translate service to proceed.": "Günlük çeviri kotası aşıldı. Devam etmek için başka bir çeviri hizmeti seçin."
+  "Daily translation quota reached. Select another translate service to proceed.": "Günlük çeviri kotası aşıldı. Devam etmek için başka bir çeviri hizmeti seçin.",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: {{count}} ses",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} sesler"
 }
@@ -353,5 +353,9 @@
  "Quota Exceeded": "Перевищено ліміт",
  "{{percentage}}% of Daily Translation Characters Used.": "Використано {{percentage}}% символів перекладу за день.",
  "Translation Characters": "Символи перекладу",
-  "Daily translation quota reached. Select another translate service to proceed.": "Досягнуто добового ліміту перекладу. Виберіть іншу службу перекладу для продовження."
+  "Daily translation quota reached. Select another translate service to proceed.": "Досягнуто добового ліміту перекладу. Виберіть іншу службу перекладу для продовження.",
+  "{{engine}}: {{count}} voices_one": "{{engine}}: {{count}} голос",
+  "{{engine}}: {{count}} voices_few": "{{engine}}: {{count}} голоси",
+  "{{engine}}: {{count}} voices_many": "{{engine}}: {{count}} голосів",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} голоси"
 }
@@ -347,5 +347,6 @@
  "Quota Exceeded": "Vượt hạn mức",
  "{{percentage}}% of Daily Translation Characters Used.": "{{percentage}}% số ký tự dịch hàng ngày đã sử dụng.",
  "Translation Characters": "Những ký tự dịch",
-  "Daily translation quota reached. Select another translate service to proceed.": "Đã đạt hạn mức dịch hàng ngày. Chọn dịch vụ dịch khác để tiếp tục."
+  "Daily translation quota reached. Select another translate service to proceed.": "Đã đạt hạn mức dịch hàng ngày. Chọn dịch vụ dịch khác để tiếp tục.",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} giọng"
 }
@@ -347,5 +347,6 @@
  "Quota Exceeded": "额度不足",
  "{{percentage}}% of Daily Translation Characters Used.": "已使用 {{percentage}}% 的每日翻译字符数",
  "Translation Characters": "翻译字数",
-  "Daily translation quota reached. Select another translate service to proceed.": "已达到每日翻译配额，请选择其他翻译服务继续。"
+  "Daily translation quota reached. Select another translate service to proceed.": "已达到每日翻译配额，请选择其他翻译服务继续。",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} 种声音"
 }
@@ -347,5 +347,6 @@
  "Quota Exceeded": "額度不足",
  "{{percentage}}% of Daily Translation Characters Used.": "已使用 {{percentage}}% 的每日翻譯字數",
  "Translation Characters": "翻譯字數",
-  "Daily translation quota reached. Select another translate service to proceed.": "已達到每日翻譯字數限制，請選擇其他翻譯服務繼續。"
+  "Daily translation quota reached. Select another translate service to proceed.": "已達到每日翻譯字數限制，請選擇其他翻譯服務繼續。",
+  "{{engine}}: {{count}} voices_other": "{{engine}}: {{count}} 種語音"
 }
@@ -1,4 +1,4 @@
-package com.readest.native-tts
+package com.readest.native_tts

 import androidx.test.platform.app.InstrumentationRegistry
 import androidx.test.ext.junit.runners.AndroidJUnit4
@@ -19,6 +19,6 @@ class ExampleInstrumentedTest {
    fun useAppContext() {
        // Context of the app under test.
        val appContext = InstrumentationRegistry.getInstrumentation().targetContext
-        assertEquals("com.readest.native-tts", appContext.packageName)
+        assertEquals("com.readest.native_tts", appContext.packageName)
    }
 }
@@ -20,10 +20,10 @@ import kotlinx.coroutines.flow.Flow
 import kotlinx.coroutines.flow.flow
 import org.json.JSONArray
 import org.json.JSONObject
-import java.util.*
 import java.util.concurrent.ConcurrentHashMap
 import java.util.concurrent.atomic.AtomicBoolean
 import java.util.concurrent.atomic.AtomicReference
+import java.util.*

 data class TTSVoiceData(
    val id: String,
@@ -38,11 +38,26 @@ data class TTSMessageEvent(
    val mark: String? = null
 )

-enum class TTSGranularity(val value: String) {
-    WORD("word"),
-    SENTENCE("sentence"),
-    PARAGRAPH("paragraph")
-}
+@InvokeArg
+class SpeakArgs(
+    val text: String? = "",
+    val preload: Boolean? = false
+)
+
+@InvokeArg
+class SetRateArgs(
+    val rate: Float? = 1.0f
+)
+
+@InvokeArg
+class SetPitchArgs(
+    val pitch: Float? = 1.0f
+)
+
+@InvokeArg
+class SetVoiceArgs(
+    val voice: String? = null
+)

@TauriPlugin
 class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
@@ -56,12 +71,9 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
    private var isInitialized = AtomicBoolean(false)
    private var isPaused = AtomicBoolean(false)
    private var isSpeaking = AtomicBoolean(false)
-    private var currentVoiceId = AtomicReference<String>("")
-    private var currentLang = AtomicReference<String>("en-US")
    private var currentRate = AtomicReference<Float>(1.0f)
    private var currentPitch = AtomicReference<Float>(1.0f)
    
-    // Event channels for each speaking session
    private val eventChannels = ConcurrentHashMap<String, Channel<TTSMessageEvent>>()
    private val speakingJobs = ConcurrentHashMap<String, Job>()
    
@@ -90,16 +102,19 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
                    TextToSpeech.SUCCESS -> {
                        setupTTSListener()
                        isInitialized.set(true)
+                        @OptIn(kotlinx.coroutines.ExperimentalCoroutinesApi::class)
                        continuation.resume(true) {}
                    }
                    else -> {
                        Log.e(TAG, "TTS initialization failed with status: $status")
+                        @OptIn(kotlinx.coroutines.ExperimentalCoroutinesApi::class)
                        continuation.resume(false) {}
                    }
                }
            }
        } catch (e: Exception) {
            Log.e(TAG, "Exception during TTS initialization", e)
+            @OptIn(kotlinx.coroutines.ExperimentalCoroutinesApi::class)
            continuation.resume(false) {}
        }
    }
@@ -120,7 +135,8 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
                    closeEventChannel(id)
                }
            }
-            
+
+            @Deprecated("deprecated in API level 21")
            override fun onError(utteranceId: String?) {
                utteranceId?.let { id ->
                    isSpeaking.set(false)
@@ -129,6 +145,14 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
                }
            }
            
+            override fun onError(utteranceId: String?, errorCode: Int) {
+                utteranceId?.let { id ->
+                    isSpeaking.set(false)
+                    sendEvent(id, TTSMessageEvent("error", "TTS playback error:$errorCode"))
+                    closeEventChannel(id)
+                }
+            }
+            
            override fun onRangeStart(utteranceId: String?, start: Int, end: Int, frame: Int) {
                utteranceId?.let { id ->
                    sendEvent(id, TTSMessageEvent("boundary", "range", "pos:$start-$end"))
@@ -140,9 +164,10 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
    @Command
    fun speak(invoke: Invoke) {
        val args = invoke.parseArgs(SpeakArgs::class.java)
+        val text = args.text ?: ""
        
-        if (!isInitialized.get()) {
-            invoke.reject("TTS not initialized")
+        if (text.isEmpty()) {
+            invoke.reject("Text cannot be empty")
            return
        }
        
@@ -154,11 +179,10 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
                eventChannels[utteranceId] = eventChannel
                
                val speakJob = launch {
-                    speakText(args.ssml, utteranceId, args.preload ?: false)
+                    speakText(text, utteranceId, args.preload ?: false)
                }
                speakingJobs[utteranceId] = speakJob
                
-                // Return the utterance ID so frontend can listen to events
                val result = JSObject().apply {
                    put("utteranceId", utteranceId)
                }
@@ -174,12 +198,9 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
        }
    }
    
-    private suspend fun speakText(ssml: String, utteranceId: String, preload: Boolean) {
+    private suspend fun speakText(text: String, utteranceId: String, preload: Boolean) {
        withContext(Dispatchers.Main) {
            try {
-                // Parse SSML and extract text
-                val text = parseSSML(ssml)
-
                textToSpeech?.apply {
                    setSpeechRate(currentRate.get())
                    setPitch(currentPitch.get())
@@ -205,18 +226,9 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
        }
    }
    
-    private fun parseSSML(ssml: String): String {
-        // Simple SSML parsing - extract text content
-        return ssml
-            .replace(Regex("<[^>]*>"), " ")
-            .replace(Regex("\\s+"), " ")
-            .trim()
-    }
-    
    private fun startEventStream(utteranceId: String) {
        coroutineScope.launch {
            val channel = eventChannels[utteranceId] ?: return@launch
-            
            try {
                for (event in channel) {
                    val eventData = JSObject().apply {
@@ -225,8 +237,6 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
                        event.message?.let { put("message", it) }
                        event.mark?.let { put("mark", it) }
                    }
-                    
-                    // Send event to frontend via Tauri event system
                    trigger(CHANNEL_NAME, eventData)
                }
            } catch (e: Exception) {
@@ -266,7 +276,6 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
    
    @Command
    fun resume(invoke: Invoke) {
-        // Android TTS doesn't have native resume, so we'll need to track state
        try {
            isPaused.set(false)
            invoke.resolve()
@@ -281,8 +290,6 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
            if (textToSpeech?.stop() == TextToSpeech.SUCCESS) {
                isSpeaking.set(false)
                isPaused.set(false)
-                
-                // Cancel all active speaking jobs and close channels
                speakingJobs.values.forEach { it.cancel() }
                eventChannels.values.forEach { it.close() }
                speakingJobs.clear()
@@ -297,29 +304,6 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
        }
    }
    
-    @Command
-    fun set_primary_lang(invoke: Invoke) {
-        val args = invoke.parseArgs(SetLangArgs::class.java)
-        try {
-            val locale = Locale.forLanguageTag(args.lang)
-            val result = textToSpeech?.setLanguage(locale)
-            
-            when (result) {
-                TextToSpeech.LANG_AVAILABLE,
-                TextToSpeech.LANG_COUNTRY_AVAILABLE,
-                TextToSpeech.LANG_COUNTRY_VAR_AVAILABLE -> {
-                    currentLang.set(args.lang)
-                    invoke.resolve()
-                }
-                else -> {
-                    invoke.reject("Language not supported: ${args.lang}")
-                }
-            }
-        } catch (e: Exception) {
-            invoke.reject("Exception setting language: ${e.message}")
-        }
-    }
-    
    @Command
    fun set_rate(invoke: Invoke) {
        val args = invoke.parseArgs(SetRateArgs::class.java)
@@ -352,7 +336,6 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
            if (targetVoice != null) {
                val result = textToSpeech?.setVoice(targetVoice)
                if (result == TextToSpeech.SUCCESS) {
-                    currentVoiceId.set(args.voice)
                    invoke.resolve()
                } else {
                    invoke.reject("Failed to set voice: ${args.voice}")
@@ -386,68 +369,6 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
        }
    }
    
-    @Command
-    fun get_voices(invoke: Invoke) {
-        val args = invoke.parseArgs(GetVoicesArgs::class.java)
-        try {
-            val locale = Locale.forLanguageTag(args.lang)
-            val voices = textToSpeech?.voices?.filter { voice ->
-                voice.locale.language == locale.language
-            }?.map { voice ->
-                JSObject().apply {
-                    put("id", voice.name)
-                    put("name", voice.name)
-                    put("lang", voice.locale.toLanguageTag())
-                    put("disabled", false)
-                }
-            } ?: emptyList()
-            
-            val result = JSObject().apply {
-                put("voices", JSONArray(voices))
-            }
-            invoke.resolve(result)
-        } catch (e: Exception) {
-            invoke.reject("Exception getting voices for language: ${e.message}")
-        }
-    }
-    
-    @Command
-    fun get_granularities(invoke: Invoke) {
-        try {
-            val granularities = TTSGranularity.values().map { it.value }
-            val result = JSObject().apply {
-                put("granularities", JSONArray(granularities))
-            }
-            invoke.resolve(result)
-        } catch (e: Exception) {
-            invoke.reject("Exception getting granularities: ${e.message}")
-        }
-    }
-    
-    @Command
-    fun get_voice_id(invoke: Invoke) {
-        try {
-            val result = JSObject().apply {
-                put("voiceId", currentVoiceId.get())
-            }
-            invoke.resolve(result)
-        } catch (e: Exception) {
-            invoke.reject("Exception getting voice ID: ${e.message}")
-        }
-    }
-    
-    @Command
-    fun get_speaking_lang(invoke: Invoke) {
-        try {
-            val result = JSObject().apply {
-                put("lang", currentLang.get())
-            }
-            invoke.resolve(result)
-        } catch (e: Exception) {
-            invoke.reject("Exception getting speaking language: ${e.message}")
-        }
-    }
-    
    fun destroy() {
        coroutineScope.cancel()
        textToSpeech?.shutdown()
@@ -457,29 +378,3 @@ class NativeTTSPlugin(private val activity: Activity) : Plugin(activity) {
        speakingJobs.clear()
    }
 }
-
-// Data classes for command arguments
-data class SpeakArgs(
-    val ssml: String,
-    val preload: Boolean? = false
-)
-
-data class SetLangArgs(
-    val lang: String
-)
-
-data class SetRateArgs(
-    val rate: Float
-)
-
-data class SetPitchArgs(
-    val pitch: Float
-)
-
-data class SetVoiceArgs(
-    val voice: String
-)
-
-data class GetVoicesArgs(
-    val lang: String
-)
@@ -7,12 +7,9 @@ const COMMANDS: &[&str] = &[
    "set_rate",
    "set_pitch",
    "set_voice",
-    "set_primary_lang",
-    "get_voices",
-    "get_voice_id",
    "get_all_voices",
-    "get_granularities",
-    "get_speaking_lang",
+    "registerListener",
+    "remove_listener",
 ];

 fn main() {
@@ -1,13 +0,0 @@
-# Automatically generated - DO NOT EDIT!
-
-"$schema" = "../../schemas/schema.json"
-
-[[permission]]
-identifier = "allow-get-granularities"
-description = "Enables the get_granularities command without any pre-configured scope."
-commands.allow = ["get_granularities"]
-
-[[permission]]
-identifier = "deny-get-granularities"
-description = "Denies the get_granularities command without any pre-configured scope."
-commands.deny = ["get_granularities"]
@@ -1,13 +0,0 @@
-# Automatically generated - DO NOT EDIT!
-
-"$schema" = "../../schemas/schema.json"
-
-[[permission]]
-identifier = "allow-get-speaking-lang"
-description = "Enables the get_speaking_lang command without any pre-configured scope."
-commands.allow = ["get_speaking_lang"]
-
-[[permission]]
-identifier = "deny-get-speaking-lang"
-description = "Denies the get_speaking_lang command without any pre-configured scope."
-commands.deny = ["get_speaking_lang"]
@@ -1,13 +0,0 @@
-# Automatically generated - DO NOT EDIT!
-
-"$schema" = "../../schemas/schema.json"
-
-[[permission]]
-identifier = "allow-get-voice-id"
-description = "Enables the get_voice_id command without any pre-configured scope."
-commands.allow = ["get_voice_id"]
-
-[[permission]]
-identifier = "deny-get-voice-id"
-description = "Denies the get_voice_id command without any pre-configured scope."
-commands.deny = ["get_voice_id"]
@@ -1,13 +0,0 @@
-# Automatically generated - DO NOT EDIT!
-
-"$schema" = "../../schemas/schema.json"
-
-[[permission]]
-identifier = "allow-get-voices"
-description = "Enables the get_voices command without any pre-configured scope."
-commands.allow = ["get_voices"]
-
-[[permission]]
-identifier = "deny-get-voices"
-description = "Denies the get_voices command without any pre-configured scope."
-commands.deny = ["get_voices"]
@@ -0,0 +1,13 @@
+# Automatically generated - DO NOT EDIT!
+
+"$schema" = "../../schemas/schema.json"
+
+[[permission]]
+identifier = "allow-registerListener"
+description = "Enables the registerListener command without any pre-configured scope."
+commands.allow = ["registerListener"]
+
+[[permission]]
+identifier = "deny-registerListener"
+description = "Denies the registerListener command without any pre-configured scope."
+commands.deny = ["registerListener"]
@@ -0,0 +1,13 @@
+# Automatically generated - DO NOT EDIT!
+
+"$schema" = "../../schemas/schema.json"
+
+[[permission]]
+identifier = "allow-remove-listener"
+description = "Enables the remove_listener command without any pre-configured scope."
+commands.allow = ["remove_listener"]
+
+[[permission]]
+identifier = "deny-remove-listener"
+description = "Denies the remove_listener command without any pre-configured scope."
+commands.deny = ["remove_listener"]
@@ -1,13 +0,0 @@
-# Automatically generated - DO NOT EDIT!
-
-"$schema" = "../../schemas/schema.json"
-
-[[permission]]
-identifier = "allow-set-primary-lang"
-description = "Enables the set_primary_lang command without any pre-configured scope."
-commands.allow = ["set_primary_lang"]
-
-[[permission]]
-identifier = "deny-set-primary-lang"
-description = "Denies the set_primary_lang command without any pre-configured scope."
-commands.deny = ["set_primary_lang"]
@@ -12,12 +12,9 @@ Default permissions for the plugin
 - `allow-set-rate`
 - `allow-set-pitch`
 - `allow-set-voice`
- `allow-set-primary-lang`
- `allow-get-voices`
- `allow-get-voice-id`
 - `allow-get-all-voices`
- `allow-get-granularities`
- `allow-get-speaking-lang`
+- `allow-registerListener`
+- `allow-remove-listener`

 ## Permission Table

@@ -57,110 +54,6 @@ Denies the get_all_voices command without any pre-configured scope.
 <tr>
 <td>

-`native-tts:allow-get-granularities`
-
-</td>
-<td>
-
-Enables the get_granularities command without any pre-configured scope.
-
-</td>
-</tr>
-
-<tr>
-<td>
-
-`native-tts:deny-get-granularities`
-
-</td>
-<td>
-
-Denies the get_granularities command without any pre-configured scope.
-
-</td>
-</tr>
-
-<tr>
-<td>
-
-`native-tts:allow-get-speaking-lang`
-
-</td>
-<td>
-
-Enables the get_speaking_lang command without any pre-configured scope.
-
-</td>
-</tr>
-
-<tr>
-<td>
-
-`native-tts:deny-get-speaking-lang`
-
-</td>
-<td>
-
-Denies the get_speaking_lang command without any pre-configured scope.
-
-</td>
-</tr>
-
-<tr>
-<td>
-
-`native-tts:allow-get-voice-id`
-
-</td>
-<td>
-
-Enables the get_voice_id command without any pre-configured scope.
-
-</td>
-</tr>
-
-<tr>
-<td>
-
-`native-tts:deny-get-voice-id`
-
-</td>
-<td>
-
-Denies the get_voice_id command without any pre-configured scope.
-
-</td>
-</tr>
-
-<tr>
-<td>
-
-`native-tts:allow-get-voices`
-
-</td>
-<td>
-
-Enables the get_voices command without any pre-configured scope.
-
-</td>
-</tr>
-
-<tr>
-<td>
-
-`native-tts:deny-get-voices`
-
-</td>
-<td>
-
-Denies the get_voices command without any pre-configured scope.
-
-</td>
-</tr>
-
-<tr>
-<td>
-
 `native-tts:allow-init`

 </td>
@@ -213,6 +106,58 @@ Denies the pause command without any pre-configured scope.
 <tr>
 <td>

+`native-tts:allow-registerListener`
+
+</td>
+<td>
+
+Enables the registerListener command without any pre-configured scope.
+
+</td>
+</tr>
+
+<tr>
+<td>
+
+`native-tts:deny-registerListener`
+
+</td>
+<td>
+
+Denies the registerListener command without any pre-configured scope.
+
+</td>
+</tr>
+
+<tr>
+<td>
+
+`native-tts:allow-remove-listener`
+
+</td>
+<td>
+
+Enables the remove_listener command without any pre-configured scope.
+
+</td>
+</tr>
+
+<tr>
+<td>
+
+`native-tts:deny-remove-listener`
+
+</td>
+<td>
+
+Denies the remove_listener command without any pre-configured scope.
+
+</td>
+</tr>
+
+<tr>
+<td>
+
 `native-tts:allow-resume`

 </td>
@@ -265,32 +210,6 @@ Denies the set_pitch command without any pre-configured scope.
 <tr>
 <td>

-`native-tts:allow-set-primary-lang`
-
-</td>
-<td>
-
-Enables the set_primary_lang command without any pre-configured scope.
-
-</td>
-</tr>
-
-<tr>
-<td>
-
-`native-tts:deny-set-primary-lang`
-
-</td>
-<td>
-
-Denies the set_primary_lang command without any pre-configured scope.
-
-</td>
-</tr>
-
-<tr>
-<td>
-
 `native-tts:allow-set-rate`

 </td>
@@ -9,10 +9,7 @@ permissions = [
  "allow-set-rate",
  "allow-set-pitch",
  "allow-set-voice",
-  "allow-set-primary-lang",
-  "allow-get-voices",
-  "allow-get-voice-id",
  "allow-get-all-voices",
-  "allow-get-granularities",
-  "allow-get-speaking-lang",
+  "allow-registerListener",
+  "allow-remove-listener",
 ]
@@ -306,54 +306,6 @@
          "const": "deny-get-all-voices",
          "markdownDescription": "Denies the get_all_voices command without any pre-configured scope."
        },
-        {
-          "description": "Enables the get_granularities command without any pre-configured scope.",
-          "type": "string",
-          "const": "allow-get-granularities",
-          "markdownDescription": "Enables the get_granularities command without any pre-configured scope."
-        },
-        {
-          "description": "Denies the get_granularities command without any pre-configured scope.",
-          "type": "string",
-          "const": "deny-get-granularities",
-          "markdownDescription": "Denies the get_granularities command without any pre-configured scope."
-        },
-        {
-          "description": "Enables the get_speaking_lang command without any pre-configured scope.",
-          "type": "string",
-          "const": "allow-get-speaking-lang",
-          "markdownDescription": "Enables the get_speaking_lang command without any pre-configured scope."
-        },
-        {
-          "description": "Denies the get_speaking_lang command without any pre-configured scope.",
-          "type": "string",
-          "const": "deny-get-speaking-lang",
-          "markdownDescription": "Denies the get_speaking_lang command without any pre-configured scope."
-        },
-        {
-          "description": "Enables the get_voice_id command without any pre-configured scope.",
-          "type": "string",
-          "const": "allow-get-voice-id",
-          "markdownDescription": "Enables the get_voice_id command without any pre-configured scope."
-        },
-        {
-          "description": "Denies the get_voice_id command without any pre-configured scope.",
-          "type": "string",
-          "const": "deny-get-voice-id",
-          "markdownDescription": "Denies the get_voice_id command without any pre-configured scope."
-        },
-        {
-          "description": "Enables the get_voices command without any pre-configured scope.",
-          "type": "string",
-          "const": "allow-get-voices",
-          "markdownDescription": "Enables the get_voices command without any pre-configured scope."
-        },
-        {
-          "description": "Denies the get_voices command without any pre-configured scope.",
-          "type": "string",
-          "const": "deny-get-voices",
-          "markdownDescription": "Denies the get_voices command without any pre-configured scope."
-        },
        {
          "description": "Enables the init command without any pre-configured scope.",
          "type": "string",
@@ -378,6 +330,30 @@
          "const": "deny-pause",
          "markdownDescription": "Denies the pause command without any pre-configured scope."
        },
+        {
+          "description": "Enables the registerListener command without any pre-configured scope.",
+          "type": "string",
+          "const": "allow-registerListener",
+          "markdownDescription": "Enables the registerListener command without any pre-configured scope."
+        },
+        {
+          "description": "Denies the registerListener command without any pre-configured scope.",
+          "type": "string",
+          "const": "deny-registerListener",
+          "markdownDescription": "Denies the registerListener command without any pre-configured scope."
+        },
+        {
+          "description": "Enables the remove_listener command without any pre-configured scope.",
+          "type": "string",
+          "const": "allow-remove-listener",
+          "markdownDescription": "Enables the remove_listener command without any pre-configured scope."
+        },
+        {
+          "description": "Denies the remove_listener command without any pre-configured scope.",
+          "type": "string",
+          "const": "deny-remove-listener",
+          "markdownDescription": "Denies the remove_listener command without any pre-configured scope."
+        },
        {
          "description": "Enables the resume command without any pre-configured scope.",
          "type": "string",
@@ -402,18 +378,6 @@
          "const": "deny-set-pitch",
          "markdownDescription": "Denies the set_pitch command without any pre-configured scope."
        },
-        {
-          "description": "Enables the set_primary_lang command without any pre-configured scope.",
-          "type": "string",
-          "const": "allow-set-primary-lang",
-          "markdownDescription": "Enables the set_primary_lang command without any pre-configured scope."
-        },
-        {
-          "description": "Denies the set_primary_lang command without any pre-configured scope.",
-          "type": "string",
-          "const": "deny-set-primary-lang",
-          "markdownDescription": "Denies the set_primary_lang command without any pre-configured scope."
-        },
        {
          "description": "Enables the set_rate command without any pre-configured scope.",
          "type": "string",
@@ -463,10 +427,10 @@
          "markdownDescription": "Denies the stop command without any pre-configured scope."
        },
        {
-          "description": "Default permissions for the plugin\n#### This default permission set includes:\n\n- `allow-init`\n- `allow-speak`\n- `allow-stop`\n- `allow-pause`\n- `allow-resume`\n- `allow-set-rate`\n- `allow-set-pitch`\n- `allow-set-voice`\n- `allow-set-primary-lang`\n- `allow-get-voices`\n- `allow-get-voice-id`\n- `allow-get-all-voices`\n- `allow-get-granularities`\n- `allow-get-speaking-lang`",
+          "description": "Default permissions for the plugin\n#### This default permission set includes:\n\n- `allow-init`\n- `allow-speak`\n- `allow-stop`\n- `allow-pause`\n- `allow-resume`\n- `allow-set-rate`\n- `allow-set-pitch`\n- `allow-set-voice`\n- `allow-get-all-voices`\n- `allow-registerListener`\n- `allow-remove-listener`",
          "type": "string",
          "const": "default",
-          "markdownDescription": "Default permissions for the plugin\n#### This default permission set includes:\n\n- `allow-init`\n- `allow-speak`\n- `allow-stop`\n- `allow-pause`\n- `allow-resume`\n- `allow-set-rate`\n- `allow-set-pitch`\n- `allow-set-voice`\n- `allow-set-primary-lang`\n- `allow-get-voices`\n- `allow-get-voice-id`\n- `allow-get-all-voices`\n- `allow-get-granularities`\n- `allow-get-speaking-lang`"
+          "markdownDescription": "Default permissions for the plugin\n#### This default permission set includes:\n\n- `allow-init`\n- `allow-speak`\n- `allow-stop`\n- `allow-pause`\n- `allow-resume`\n- `allow-set-rate`\n- `allow-set-pitch`\n- `allow-set-voice`\n- `allow-get-all-voices`\n- `allow-registerListener`\n- `allow-remove-listener`"
        }
      ]
    }
@@ -5,74 +5,49 @@ use crate::NativeTtsExt;
 use crate::Result;

 #[command]
-pub async fn init<R: Runtime>(app: AppHandle<R>) -> Result<bool> {
+pub(crate) async fn init<R: Runtime>(app: AppHandle<R>) -> Result<InitResponse> {
    app.native_tts().init()
 }

 #[command]
-pub async fn speak<R: Runtime>(app: AppHandle<R>, args: SpeakArgs) -> Result<String> {
-    app.native_tts().speak(args)
+pub(crate) async fn speak<R: Runtime>(
+    app: AppHandle<R>,
+    payload: SpeakArgs,
+) -> Result<SpeakResponse> {
+    app.native_tts().speak(payload)
 }

 #[command]
-pub async fn pause<R: Runtime>(app: AppHandle<R>) -> Result<()> {
+pub(crate) async fn pause<R: Runtime>(app: AppHandle<R>) -> Result<()> {
    app.native_tts().pause()
 }

 #[command]
-pub async fn resume<R: Runtime>(app: AppHandle<R>) -> Result<()> {
+pub(crate) async fn resume<R: Runtime>(app: AppHandle<R>) -> Result<()> {
    app.native_tts().resume()
 }

 #[command]
-pub async fn stop<R: Runtime>(app: AppHandle<R>) -> Result<()> {
+pub(crate) async fn stop<R: Runtime>(app: AppHandle<R>) -> Result<()> {
    app.native_tts().stop()
 }

 #[command]
-pub async fn set_primary_lang<R: Runtime>(app: AppHandle<R>, args: SetLangArgs) -> Result<()> {
-    app.native_tts().set_primary_lang(args)
+pub(crate) async fn set_rate<R: Runtime>(app: AppHandle<R>, payload: SetRateArgs) -> Result<()> {
+    app.native_tts().set_rate(payload)
 }

 #[command]
-pub async fn set_rate<R: Runtime>(app: AppHandle<R>, args: SetRateArgs) -> Result<()> {
-    app.native_tts().set_rate(args)
+pub(crate) async fn set_pitch<R: Runtime>(app: AppHandle<R>, payload: SetPitchArgs) -> Result<()> {
+    app.native_tts().set_pitch(payload)
 }

 #[command]
-pub async fn set_pitch<R: Runtime>(app: AppHandle<R>, args: SetPitchArgs) -> Result<()> {
-    app.native_tts().set_pitch(args)
+pub(crate) async fn set_voice<R: Runtime>(app: AppHandle<R>, payload: SetVoiceArgs) -> Result<()> {
+    app.native_tts().set_voice(payload)
 }

 #[command]
-pub async fn set_voice<R: Runtime>(app: AppHandle<R>, args: SetVoiceArgs) -> Result<()> {
-    app.native_tts().set_voice(args)
-}
-
-#[command]
-pub async fn get_all_voices<R: Runtime>(app: AppHandle<R>) -> Result<Vec<TTSVoice>> {
+pub(crate) async fn get_all_voices<R: Runtime>(app: AppHandle<R>) -> Result<GetVoicesResponse> {
    app.native_tts().get_all_voices()
 }
-
-#[command]
-pub async fn get_voices<R: Runtime>(
-    app: AppHandle<R>,
-    args: GetVoicesArgs,
-) -> Result<Vec<TTSVoice>> {
-    app.native_tts().get_voices(args)
-}
-
-#[command]
-pub async fn get_granularities<R: Runtime>(app: AppHandle<R>) -> Result<Vec<TTSGranularity>> {
-    app.native_tts().get_granularities()
-}
-
-#[command]
-pub async fn get_voice_id<R: Runtime>(app: AppHandle<R>) -> Result<String> {
-    app.native_tts().get_voice_id()
-}
-
-#[command]
-pub async fn get_speaking_lang<R: Runtime>(app: AppHandle<R>) -> Result<String> {
-    app.native_tts().get_speaking_lang()
-}
@@ -14,10 +14,10 @@ pub fn init<R: Runtime, C: DeserializeOwned>(
 pub struct NativeTts<R: Runtime>(AppHandle<R>);

 impl<R: Runtime> NativeTts<R> {
-    pub fn init(&self) -> crate::Result<bool> {
+    pub fn init(&self) -> crate::Result<InitResponse> {
        Err(crate::Error::UnsupportedPlatformError)
    }
-    pub fn speak(&self, _args: SpeakArgs) -> crate::Result<String> {
+    pub fn speak(&self, _args: SpeakArgs) -> crate::Result<SpeakResponse> {
        Err(crate::Error::UnsupportedPlatformError)
    }
    pub fn pause(&self) -> crate::Result<()> {
@@ -29,9 +29,6 @@ impl<R: Runtime> NativeTts<R> {
    pub fn stop(&self) -> crate::Result<()> {
        Err(crate::Error::UnsupportedPlatformError)
    }
-    pub fn set_primary_lang(&self, _args: SetLangArgs) -> crate::Result<()> {
-        Err(crate::Error::UnsupportedPlatformError)
-    }
    pub fn set_rate(&self, _args: SetRateArgs) -> crate::Result<()> {
        Err(crate::Error::UnsupportedPlatformError)
    }
@@ -41,19 +38,7 @@ impl<R: Runtime> NativeTts<R> {
    pub fn set_voice(&self, _args: SetVoiceArgs) -> crate::Result<()> {
        Err(crate::Error::UnsupportedPlatformError)
    }
-    pub fn get_all_voices(&self) -> crate::Result<Vec<TTSVoice>> {
-        Err(crate::Error::UnsupportedPlatformError)
-    }
-    pub fn get_voices(&self, _args: GetVoicesArgs) -> crate::Result<Vec<TTSVoice>> {
-        Err(crate::Error::UnsupportedPlatformError)
-    }
-    pub fn get_granularities(&self) -> crate::Result<Vec<TTSGranularity>> {
-        Err(crate::Error::UnsupportedPlatformError)
-    }
-    pub fn get_voice_id(&self) -> crate::Result<String> {
-        Err(crate::Error::UnsupportedPlatformError)
-    }
-    pub fn get_speaking_lang(&self) -> crate::Result<String> {
+    pub fn get_all_voices(&self) -> crate::Result<GetVoicesResponse> {
        Err(crate::Error::UnsupportedPlatformError)
    }
 }
@@ -44,12 +44,7 @@ pub fn init<R: Runtime>() -> TauriPlugin<R> {
            commands::set_rate,
            commands::set_pitch,
            commands::set_voice,
-            commands::set_primary_lang,
-            commands::get_voices,
-            commands::get_voice_id,
            commands::get_all_voices,
-            commands::get_granularities,
-            commands::get_speaking_lang
        ])
        .setup(|app, api| {
            #[cfg(mobile)]
@@ -25,14 +25,16 @@ pub fn init<R: Runtime, C: DeserializeOwned>(
 pub struct NativeTts<R: Runtime>(PluginHandle<R>);

 impl<R: Runtime> NativeTts<R> {
-    pub fn init(&self) -> crate::Result<bool> {
+    pub fn init(&self) -> crate::Result<InitResponse> {
        self.0.run_mobile_plugin("init", ()).map_err(Into::into)
    }
 }

 impl<R: Runtime> NativeTts<R> {
-    pub fn speak(&self, args: SpeakArgs) -> crate::Result<String> {
-        self.0.run_mobile_plugin("speak", args).map_err(Into::into)
+    pub fn speak(&self, payload: SpeakArgs) -> crate::Result<SpeakResponse> {
+        self.0
+            .run_mobile_plugin("speak", payload)
+            .map_err(Into::into)
    }
 }

@@ -55,73 +57,33 @@ impl<R: Runtime> NativeTts<R> {
 }

 impl<R: Runtime> NativeTts<R> {
-    pub fn set_primary_lang(&self, args: SetLangArgs) -> crate::Result<()> {
+    pub fn set_rate(&self, payload: SetRateArgs) -> crate::Result<()> {
        self.0
-            .run_mobile_plugin("set_primary_lang", args)
+            .run_mobile_plugin("set_rate", payload)
            .map_err(Into::into)
    }
 }

 impl<R: Runtime> NativeTts<R> {
-    pub fn set_rate(&self, args: SetRateArgs) -> crate::Result<()> {
+    pub fn set_pitch(&self, payload: SetPitchArgs) -> crate::Result<()> {
        self.0
-            .run_mobile_plugin("set_rate", args)
+            .run_mobile_plugin("set_pitch", payload)
            .map_err(Into::into)
    }
 }

 impl<R: Runtime> NativeTts<R> {
-    pub fn set_pitch(&self, args: SetPitchArgs) -> crate::Result<()> {
+    pub fn set_voice(&self, payload: SetVoiceArgs) -> crate::Result<()> {
        self.0
-            .run_mobile_plugin("set_pitch", args)
+            .run_mobile_plugin("set_voice", payload)
            .map_err(Into::into)
    }
 }

 impl<R: Runtime> NativeTts<R> {
-    pub fn set_voice(&self, args: SetVoiceArgs) -> crate::Result<()> {
-        self.0
-            .run_mobile_plugin("set_voice", args)
-            .map_err(Into::into)
-    }
-}
-
-impl<R: Runtime> NativeTts<R> {
-    pub fn get_voices(&self, args: GetVoicesArgs) -> crate::Result<Vec<TTSVoice>> {
-        self.0
-            .run_mobile_plugin("get_voices", args)
-            .map_err(Into::into)
-    }
-}
-
-impl<R: Runtime> NativeTts<R> {
-    pub fn get_all_voices(&self) -> crate::Result<Vec<TTSVoice>> {
+    pub fn get_all_voices(&self) -> crate::Result<GetVoicesResponse> {
        self.0
            .run_mobile_plugin("get_all_voices", ())
            .map_err(Into::into)
    }
 }
-
-impl<R: Runtime> NativeTts<R> {
-    pub fn get_granularities(&self) -> crate::Result<Vec<TTSGranularity>> {
-        self.0
-            .run_mobile_plugin("get_granularities", ())
-            .map_err(Into::into)
-    }
-}
-
-impl<R: Runtime> NativeTts<R> {
-    pub fn get_voice_id(&self) -> crate::Result<String> {
-        self.0
-            .run_mobile_plugin("get_voice_id", ())
-            .map_err(Into::into)
-    }
-}
-
-impl<R: Runtime> NativeTts<R> {
-    pub fn get_speaking_lang(&self) -> crate::Result<String> {
-        self.0
-            .run_mobile_plugin("get_speaking_lang", ())
-            .map_err(Into::into)
-    }
-}
@@ -1,6 +1,7 @@
 use serde::{Deserialize, Serialize};

 #[derive(Debug, Clone, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
 pub struct TTSVoice {
    pub id: String,
    pub name: String,
@@ -10,50 +11,53 @@ pub struct TTSVoice {
 }

 #[derive(Debug, Clone, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
 pub struct TTSMessageEvent {
    pub code: String, // 'boundary' | 'error' | 'end'
    pub message: Option<String>,
    pub mark: Option<String>,
 }

-#[derive(Debug, Clone, Serialize, Deserialize)]
-pub enum TTSGranularity {
-    #[serde(rename = "word")]
-    Word,
-    #[serde(rename = "sentence")]
-    Sentence,
-    #[serde(rename = "paragraph")]
-    Paragraph,
+#[derive(Debug, Deserialize, Serialize)]
+#[serde(rename_all = "camelCase")]
+pub struct InitResponse {
+    pub success: bool,
 }

 #[derive(Debug, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
 pub struct SpeakArgs {
-    pub ssml: String,
+    pub text: String,
    #[serde(default)]
    pub preload: bool,
 }

 #[derive(Debug, Serialize, Deserialize)]
-pub struct SetLangArgs {
-    pub lang: String,
+#[serde(rename_all = "camelCase")]
+pub struct SpeakResponse {
+    pub utterance_id: String,
 }

 #[derive(Debug, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
 pub struct SetRateArgs {
    pub rate: f32,
 }

 #[derive(Debug, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
 pub struct SetPitchArgs {
    pub pitch: f32,
 }

 #[derive(Debug, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
 pub struct SetVoiceArgs {
    pub voice: String,
 }

 #[derive(Debug, Serialize, Deserialize)]
-pub struct GetVoicesArgs {
-    pub lang: String,
+#[serde(rename_all = "camelCase")]
+pub struct GetVoicesResponse {
+    pub voices: Vec<TTSVoice>,
 }
@@ -134,6 +134,7 @@ pub fn run() {
        .plugin(tauri_plugin_os::init())
        .plugin(tauri_plugin_dialog::init())
        .plugin(tauri_plugin_native_bridge::init())
+        .plugin(tauri_plugin_native_tts::init())
        .plugin(tauri_plugin_fs::init());

    #[cfg(desktop)]
@@ -90,7 +90,7 @@ const TTSControl = () => {
  useEffect(() => {
    return () => {
      if (ttsControllerRef.current) {
-        ttsControllerRef.current.kill();
+        ttsControllerRef.current.shutdown();
        ttsControllerRef.current = null;
      }
    };
@@ -202,7 +202,7 @@ const TTSControl = () => {
        unblockAudio();
      }
      setTtsClientsInitialized(false);
-      const ttsController = new TTSController(view);
+      const ttsController = new TTSController(appService, view);
      await ttsController.init();
      await ttsController.initViewTTS();
      const ssml = view.tts?.from(ttsFromRange);
@@ -3,7 +3,7 @@ import React, { useState, ChangeEvent, useEffect } from 'react';
 import { MdPlayCircle, MdPauseCircle, MdFastRewind, MdFastForward, MdAlarm } from 'react-icons/md';
 import { RiVoiceAiFill } from 'react-icons/ri';
 import { MdCheck } from 'react-icons/md';
-import { TTSVoice } from '@/services/tts';
+import { TTSVoicesGroup } from '@/services/tts';
 import { useEnv } from '@/context/EnvContext';
 import { useReaderStore } from '@/store/readerStore';
 import { TranslationFunc, useTranslation } from '@/hooks/useTranslation';
@@ -20,7 +20,7 @@ type TTSPanelProps = {
  onBackward: () => void;
  onForward: () => void;
  onSetRate: (rate: number) => void;
-  onGetVoices: (lang: string) => Promise<TTSVoice[]>;
+  onGetVoices: (lang: string) => Promise<TTSVoicesGroup[]>;
  onSetVoice: (voice: string, lang: string) => void;
  onGetVoiceId: () => string;
  onSelectTimeout: (bookKey: string, value: number) => void;
@@ -119,7 +119,7 @@ const TTSPanel = ({
  const { settings, setSettings, saveSettings } = useSettingsStore();
  const viewSettings = getViewSettings(bookKey);

-  const [voices, setVoices] = useState<TTSVoice[]>([]);
+  const [voiceGroups, setVoiceGroups] = useState<TTSVoicesGroup[]>([]);
  const [rate, setRate] = useState(viewSettings?.ttsRate ?? 1.0);
  const [selectedVoice, setSelectedVoice] = useState(viewSettings?.ttsVoice ?? '');

@@ -177,8 +177,8 @@ const TTSPanel = ({

  useEffect(() => {
    const fetchVoices = async () => {
-      const voices = await onGetVoices(ttsLang);
-      setVoices(voices);
+      const voiceGroups = await onGetVoices(ttsLang);
+      setVoiceGroups(voiceGroups);
    };
    fetchVoices();
    // eslint-disable-next-line react-hooks/exhaustive-deps
@@ -261,7 +261,12 @@ const TTSPanel = ({
                onClick={() => onSelectTimeout(bookKey, option.value)}
              >
                <div className='flex items-center px-2'>
-                  <span style={{ minWidth: `${defaultIconSize}px` }}>
+                  <span
+                    style={{
+                      width: `${defaultIconSize}px`,
+                      height: `${defaultIconSize}px`,
+                    }}
+                  >
                    {timeoutOption === option.value && <MdCheck className='text-base-content' />}
                  </span>
                  <span className={clsx('text-base sm:text-sm')}>{option.label}</span>
@@ -282,21 +287,52 @@ const TTSPanel = ({
              'mt-4 inline max-h-96 w-[250px] overflow-y-scroll',
            )}
          >
-            {voices.map((voice, index) => (
-              <li
-                key={`${index}-${voice.id}`}
-                onClick={() => !voice.disabled && handleSelectVoice(voice.id, voice.lang)}
-              >
-                <div className='flex items-center px-2'>
-                  <span style={{ minWidth: `${defaultIconSize}px` }}>
-                    {selectedVoice === voice.id && <MdCheck className='text-base-content' />}
-                  </span>
-                  <span className={clsx('text-base sm:text-sm', voice.disabled && 'text-gray-400')}>
-                    {voice.name}
-                  </span>
-                </div>
-              </li>
-            ))}
+            {voiceGroups.map((voiceGroup, index) => {
+              return (
+                voiceGroup.voices.length > 0 && (
+                  <div key={voiceGroup.id} className=''>
+                    <div className='flex items-center gap-2 px-2 py-1'>
+                      <span
+                        style={{ width: `${defaultIconSize}px`, height: `${defaultIconSize}px` }}
+                      ></span>
+                      <span className='text-sm text-gray-400 sm:text-xs'>
+                        {_('{{engine}}: {{count}} voices', {
+                          engine: _(voiceGroup.name),
+                          count: voiceGroup.voices.length,
+                        })}
+                      </span>
+                    </div>
+                    {voiceGroup.voices.map((voice, voiceIndex) => (
+                      <li
+                        key={`${index}-${voiceGroup.id}-${voiceIndex}`}
+                        onClick={() => !voice.disabled && handleSelectVoice(voice.id, voice.lang)}
+                      >
+                        <div className='flex items-center px-2'>
+                          <span
+                            style={{
+                              width: `${defaultIconSize}px`,
+                              height: `${defaultIconSize}px`,
+                            }}
+                          >
+                            {selectedVoice === voice.id && (
+                              <MdCheck className='text-base-content' />
+                            )}
+                          </span>
+                          <span
+                            className={clsx(
+                              'text-base sm:text-sm',
+                              voice.disabled && 'text-gray-400',
+                            )}
+                          >
+                            {_(voice.name)}
+                          </span>
+                        </div>
+                      </li>
+                    ))}
+                  </div>
+                )
+              );
+            })}
          </ul>
        </div>
      </div>
@@ -1,6 +1,7 @@
 import { md5 } from 'js-md5';
 import { randomMd5 } from '@/utils/misc';
 import { LRUCache } from '@/utils/lru';
+import { genSSML } from '@/utils/ssml';

 const EDGE_SPEECH_URL =
  'wss://speech.platform.bing.com/consumer/speech/synthesize/readaloud/edge/v1';
@@ -173,19 +174,6 @@ export class EdgeSpeechTTS {
      },
    });

-    const genSSML = (lang: string, text: string, voice: string, rate: number) => {
-      const cleanedText = text.replace(/^<break\b[^>]*>/i, '');
-      return `
-        <speak version="1.0" xml:lang="${lang}">
-          <voice name="${voice}">
-            <prosody rate="${rate}">
-              ${cleanedText}
-            </prosody>
-          </voice>
-        </speak>
-      `;
-    };
-
    const genSendContent = (headerObj: Record<string, string>, content: string) => {
      let header = '';
      for (const key of Object.keys(headerObj)) {
@@ -1,26 +1,28 @@
 import { getUserLocale } from '@/utils/misc';
-import { TTSClient, TTSMessageEvent, TTSVoice } from './TTSClient';
+import { TTSClient, TTSMessageEvent } from './TTSClient';
 import { EdgeSpeechTTS, EdgeTTSPayload } from '@/libs/edgeTTS';
 import { parseSSMLLang, parseSSMLMarks } from '@/utils/ssml';
-import { TTSGranularity } from '@/types/view';
 import { TTSController } from './TTSController';
 import { TTSUtils } from './TTSUtils';
+import { TTSGranularity, TTSVoice, TTSVoicesGroup } from './types';

 export class EdgeTTSClient implements TTSClient {
+  name = 'edge-tts';
+  initialized = false;
  controller?: TTSController;
+
+  #voices: TTSVoice[] = [];
  #primaryLang = 'en';
  #speakingLang = '';
+  #currentVoiceId = '';
  #rate = 1.0;
  #pitch = 1.0;
-  #voice: TTSVoice | null = null;
-  #voices: TTSVoice[] = [];
-  #edgeTTS: EdgeSpeechTTS;

+  #edgeTTS: EdgeSpeechTTS;
  #audioElement: HTMLAudioElement | null = null;
  #isPlaying = false;
  #pausedAt = 0;
  #startedAt = 0;
-  available = true;

  constructor(controller?: TTSController) {
    this.controller = controller;
@@ -37,11 +39,11 @@ export class EdgeTTSClient implements TTSClient {
        rate: 1.0,
        pitch: 1.0,
      });
-      this.available = true;
+      this.initialized = true;
    } catch {
-      this.available = false;
+      this.initialized = false;
    }
-    return this.available;
+    return this.initialized;
  }

  getPayload = (lang: string, text: string, voiceId: string) => {
@@ -49,21 +51,15 @@ export class EdgeTTSClient implements TTSClient {
  };

  getVoiceIdFromLang = async (lang: string) => {
-    let voiceId = 'en-US-AriaNeural';
-    const preferredVoiceId = TTSUtils.getPreferredVoice('edge-tts', lang);
+    const preferredVoiceId = TTSUtils.getPreferredVoice(this.name, lang);
    const preferredVoice = this.#voices.find((v) => v.id === preferredVoiceId);
-    this.#voice = preferredVoice ? preferredVoice : (await this.getVoices(lang))[0] || null;
-    if (this.#voice) {
-      voiceId = this.#voice.id;
-    }
-    return voiceId;
+    const defaultVoice = preferredVoice
+      ? preferredVoice
+      : (await this.getVoices(lang))[0]?.voices[0] || null;
+    return defaultVoice?.id || this.#currentVoiceId || 'en-US-AriaNeural';
  };

-  async *speak(
-    ssml: string,
-    signal: AbortSignal,
-    preload = false,
-  ): AsyncGenerator<TTSMessageEvent> {
+  async *speak(ssml: string, signal: AbortSignal, preload = false) {
    const { marks } = parseSSMLMarks(ssml);
    let defaultLang = parseSSMLLang(ssml) || 'en';
    if (defaultLang === 'en' && this.#primaryLang && this.#primaryLang !== 'en') {
@@ -78,6 +74,7 @@ export class EdgeTTSClient implements TTSClient {
        const { language } = mark;
        const voiceLang = language || defaultLang;
        const voiceId = await this.getVoiceIdFromLang(voiceLang);
+        this.#currentVoiceId = voiceId;
        await this.#edgeTTS
          .createAudio(this.getPayload(voiceLang, mark.text, voiceId))
          .catch((err) => {
@@ -103,7 +100,7 @@ export class EdgeTTSClient implements TTSClient {
      yield {
        code: 'end',
        message: 'Preload finished',
-      };
+      } as TTSMessageEvent;

      return;
    } else {
@@ -115,7 +112,7 @@ export class EdgeTTSClient implements TTSClient {
        yield {
          code: 'error',
          message: 'Aborted',
-        };
+        } as TTSMessageEvent;
        break;
      }
      try {
@@ -138,7 +135,7 @@ export class EdgeTTSClient implements TTSClient {
          code: 'boundary',
          message: `Start chunk: ${mark.name}`,
          mark: mark.name,
-        };
+        } as TTSMessageEvent;

        const result = await new Promise<TTSMessageEvent>((resolve) => {
          const cleanUp = () => {
@@ -180,14 +177,14 @@ export class EdgeTTSClient implements TTSClient {
          yield {
            code: 'end',
            message: `Chunk finished: ${mark.name}`,
-          };
+          } as TTSMessageEvent;
          continue;
        }
        console.log('Error:', error);
        yield {
          code: 'error',
          message: error instanceof Error ? error.message : String(error),
-        };
+        } as TTSMessageEvent;
        break;
      }

@@ -196,17 +193,19 @@ export class EdgeTTSClient implements TTSClient {
  }

  async pause() {
-    if (!this.#isPlaying || !this.#audioElement) return;
+    if (!this.#isPlaying || !this.#audioElement) return true;
    this.#pausedAt = this.#audioElement.currentTime - this.#startedAt;
    await this.#audioElement.pause();
    this.#isPlaying = false;
+    return true;
  }

  async resume() {
-    if (this.#isPlaying || !this.#audioElement) return;
+    if (this.#isPlaying || !this.#audioElement) return true;
    await this.#audioElement.play();
    this.#isPlaying = true;
    this.#startedAt = this.#audioElement.currentTime - this.#pausedAt;
+    return true;
  }

  async stop() {
@@ -231,10 +230,6 @@ export class EdgeTTSClient implements TTSClient {
    }
  }

-  setPrimaryLang(lang: string) {
-    this.#primaryLang = lang;
-  }
-
  async setRate(rate: number) {
    // The Edge TTS API uses rate in [0.5 .. 2.0].
    this.#rate = rate;
@@ -248,42 +243,36 @@ export class EdgeTTSClient implements TTSClient {
  async setVoice(voice: string) {
    const selectedVoice = this.#voices.find((v) => v.id === voice);
    if (selectedVoice) {
-      this.#voice = selectedVoice;
+      this.#currentVoiceId = selectedVoice.id;
    }
  }

  async getAllVoices(): Promise<TTSVoice[]> {
    this.#voices.forEach((voice) => {
-      voice.disabled = !this.available;
+      voice.disabled = !this.initialized;
    });
    return this.#voices;
  }

-  async getVoices(lang: string): Promise<TTSVoice[]> {
+  async getVoices(lang: string) {
    const locale = lang === 'en' ? getUserLocale(lang) || lang : lang;
    const voices = await this.getAllVoices();
-    return voices
-      .filter(
-        (v) => v.lang.startsWith(locale) || (lang === 'en' && ['en-US', 'en-GB'].includes(v.lang)),
-      )
-      .sort((a, b) => {
-        const aRegion = a.lang.split('-')[1] || '';
-        const bRegion = b.lang.split('-')[1] || '';
-        if (aRegion === bRegion) {
-          return a.name < b.name ? -1 : a.name > b.name ? 1 : 0;
-        }
-        if (aRegion === 'CN') return -1;
-        if (bRegion === 'CN') return 1;
-        if (aRegion === 'TW') return -1;
-        if (bRegion === 'TW') return 1;
-        if (aRegion === 'HK') return -1;
-        if (bRegion === 'HK') return 1;
-        if (aRegion === 'US') return -1;
-        if (bRegion === 'US') return 1;
-        if (aRegion === 'GB') return -1;
-        if (bRegion === 'GB') return 1;
-        return a.name < b.name ? -1 : a.name > b.name ? 1 : 0;
-      });
+    const filteredVoices = voices.filter(
+      (v) => v.lang.startsWith(locale) || (lang === 'en' && ['en-US', 'en-GB'].includes(v.lang)),
+    );
+
+    const voicesGroup: TTSVoicesGroup = {
+      id: 'edge-tts',
+      name: 'Edge TTS',
+      voices: filteredVoices,
+      disabled: !this.initialized || filteredVoices.length === 0,
+    };
+
+    return [voicesGroup];
+  }
+
+  setPrimaryLang(lang: string) {
+    this.#primaryLang = lang;
  }

  getGranularities(): TTSGranularity[] {
@@ -291,10 +280,16 @@ export class EdgeTTSClient implements TTSClient {
  }

  getVoiceId(): string {
-    return this.#voice?.id || '';
+    return this.#currentVoiceId;
  }

  getSpeakingLang(): string {
    return this.#speakingLang;
  }
+
+  async shutdown(): Promise<void> {
+    this.initialized = false;
+    this.#audioElement = null;
+    this.#voices = [];
+  }
 }
@@ -0,0 +1,308 @@
+import { invoke } from '@tauri-apps/api/core';
+import { addPluginListener, PluginListener } from '@tauri-apps/api/core';
+import { getUserLocale } from '@/utils/misc';
+import { parseSSMLLang, parseSSMLMarks } from '@/utils/ssml';
+import { stubTranslation as _ } from '@/utils/misc';
+import { TTSClient, TTSMessageEvent } from './TTSClient';
+import { TTSGranularity, TTSMark, TTSVoice, TTSVoicesGroup } from './types';
+import { TTSUtils } from './TTSUtils';
+import { TTSController } from './TTSController';
+
+type TTSEventPayload = {
+  utteranceId: string;
+} & TTSMessageEvent;
+
+const TTSEngines = {
+  default: 'System TTS',
+  mstrans: 'MSTrans TTS',
+  microsoft: 'Microsoft TTS',
+  edgetts: 'Edge TTS',
+  google: 'Google TTS',
+  gemini: 'Gemini TTS',
+  weread: 'WeRead TTS',
+} as Record<string, string>;
+
+export class NativeTTSClient implements TTSClient {
+  name = 'native-tts';
+  initialized = false;
+  controller?: TTSController;
+
+  #voices: TTSVoice[] = [];
+  #primaryLang = 'en';
+  #speakingLang = '';
+  #currentVoiceId = '';
+  #rate = 1.0;
+  #pitch = 1.0;
+
+  #eventListener: PluginListener | null = null;
+  #activeUtterances = new Map<
+    string,
+    {
+      eventQueue: TTSMessageEvent[];
+      resolver: ((value: IteratorResult<TTSMessageEvent>) => void) | null;
+      finished: boolean;
+    }
+  >();
+
+  constructor(controller?: TTSController) {
+    this.controller = controller;
+  }
+
+  private async setupEventListener(): Promise<void> {
+    try {
+      if (this.#eventListener) return;
+      this.#eventListener = await addPluginListener<TTSEventPayload>(
+        'native-tts',
+        'tts_events',
+        (event) => {
+          const { utteranceId, code, message, mark } = event;
+
+          const utteranceData = this.#activeUtterances.get(utteranceId);
+          if (!utteranceData) return;
+
+          const ttsEvent: TTSMessageEvent = { code, message, mark };
+          utteranceData.eventQueue.push(ttsEvent);
+          if (code === 'end' || code === 'error') {
+            utteranceData.finished = true;
+            if (utteranceData.resolver) {
+              utteranceData.resolver({ value: undefined, done: true });
+            }
+          } else if (utteranceData.resolver) {
+            utteranceData.resolver({ value: ttsEvent, done: false });
+            utteranceData.resolver = null;
+          }
+        },
+      );
+    } catch (error) {
+      console.error('Failed to setup TTS event listener:', error);
+    }
+  }
+
+  async init(): Promise<boolean> {
+    const result = await invoke<{ success: boolean }>('plugin:native-tts|init');
+    this.initialized = result.success;
+    if (this.initialized) {
+      this.setupEventListener();
+    }
+    return this.initialized;
+  }
+
+  async *speakMark(mark: TTSMark, preload: boolean, defaultLang: string, signal: AbortSignal) {
+    if (preload) {
+      yield { code: 'end', message: 'Dummy preload finished' } as TTSMessageEvent;
+      return;
+    }
+    const { language } = mark;
+    const voiceLang = language || defaultLang;
+    const voiceId = await this.getVoiceIdFromLang(voiceLang);
+    this.#currentVoiceId = voiceId;
+    this.#speakingLang = voiceLang;
+    await this.setVoice(voiceId);
+    try {
+      const result = await invoke<{ utteranceId: string }>('plugin:native-tts|speak', {
+        payload: { text: mark.text, preload },
+      });
+
+      const utteranceId = result.utteranceId;
+      this.#activeUtterances.set(utteranceId, {
+        eventQueue: [],
+        resolver: null,
+        finished: false,
+      });
+
+      const abortHandler = () => {
+        const utteranceData = this.#activeUtterances.get(utteranceId);
+        if (utteranceData && utteranceData.resolver) {
+          const error = { code: 'error', message: 'Aborted' } as TTSMessageEvent;
+          utteranceData.resolver({ value: error, done: true });
+          utteranceData.resolver = null;
+        }
+        this.stop();
+      };
+
+      signal.addEventListener('abort', abortHandler);
+
+      try {
+        while (true) {
+          const utteranceData = this.#activeUtterances.get(utteranceId);
+          if (!utteranceData) break;
+
+          if (utteranceData.eventQueue.length > 0) {
+            const event = utteranceData.eventQueue.shift()!;
+            event.mark = mark.name;
+            yield event;
+
+            if (event.code === 'end' || event.code === 'error') {
+              break;
+            }
+          } else if (utteranceData.finished) {
+            break;
+          } else {
+            const eventPromise = new Promise<TTSMessageEvent | void>((resolve) => {
+              const utteranceData = this.#activeUtterances.get(utteranceId);
+              if (!utteranceData) return resolve();
+              utteranceData.resolver = (res) =>
+                resolve(res.value?.code === 'error' ? res.value : undefined);
+            });
+            const timeoutPromise = new Promise<null>((resolve) =>
+              setTimeout(() => resolve(null), 100),
+            );
+            const result = await Promise.race([eventPromise, timeoutPromise]);
+            if (result) yield result;
+          }
+
+          if (signal.aborted) {
+            break;
+          }
+        }
+      } finally {
+        signal.removeEventListener('abort', abortHandler);
+        this.#activeUtterances.delete(utteranceId);
+      }
+    } catch (error) {
+      console.error('Failed to speak:', error);
+      throw error;
+    }
+  }
+
+  async *speak(ssml: string, signal: AbortSignal, preload: boolean = false) {
+    const { marks } = parseSSMLMarks(ssml);
+    let defaultLang = parseSSMLLang(ssml) || 'en';
+    if (defaultLang === 'en' && this.#primaryLang && this.#primaryLang !== 'en') {
+      defaultLang = this.#primaryLang;
+    }
+
+    for (const mark of marks) {
+      for await (const ev of this.speakMark(mark, preload, defaultLang, signal)) {
+        if (signal.aborted) {
+          yield { code: 'error', message: 'Aborted' } as TTSMessageEvent;
+          return;
+        }
+        yield ev;
+      }
+    }
+  }
+
+  async getVoiceIdFromLang(lang: string) {
+    const preferredVoiceId = TTSUtils.getPreferredVoice(this.name, lang);
+    const preferredVoice = this.#voices.find((v) => v.id === preferredVoiceId);
+    const defaultVoice = preferredVoice
+      ? preferredVoice
+      : (await this.getVoices(lang))[0]?.voices[0] || null;
+    return defaultVoice?.id || 'NOT_SET';
+  }
+
+  async pause() {
+    await invoke('plugin:native-tts|pause');
+    return false;
+  }
+
+  async resume() {
+    // No-op for Android TextToSpeech
+    await invoke('plugin:native-tts|resume');
+    return false;
+  }
+
+  async stop() {
+    await invoke('plugin:native-tts|stop');
+    this.#activeUtterances.clear();
+  }
+
+  async setRate(rate: number) {
+    // Power the rate to match the EdgeTTS behavior
+    this.#rate = parseFloat(Math.pow(rate, 2.5).toFixed(2));
+    await invoke('plugin:native-tts|set_rate', { payload: { rate: this.#rate } });
+  }
+
+  async setPitch(pitch: number) {
+    this.#pitch = pitch;
+    await invoke('plugin:native-tts|set_pitch', { payload: { pitch: this.#pitch } });
+  }
+
+  async setVoice(voice: string) {
+    this.#currentVoiceId = voice;
+    await invoke('plugin:native-tts|set_voice', { payload: { voice } });
+  }
+
+  async getAllVoices() {
+    if (this.#voices.length > 0) {
+      return this.#voices;
+    }
+    try {
+      const result = await invoke<{ voices: TTSVoice[] }>('plugin:native-tts|get_all_voices');
+      this.#voices = result.voices;
+      return this.#voices;
+    } catch (error) {
+      console.error('Failed to get all voices:', error);
+      return [];
+    }
+  }
+
+  async getVoices(lang: string) {
+    const locale = lang === 'en' ? getUserLocale(lang) || lang : lang;
+    const voices = await this.getAllVoices();
+    const filteredVoices = voices.filter(
+      (v) => v.lang.startsWith(locale) || (lang === 'en' && ['en-US', 'en-GB'].includes(v.lang)),
+    );
+
+    const voiceGroups = new Map<string, TTSVoice[]>();
+    filteredVoices.forEach((voice) => {
+      const { name, lang } = voice;
+      let groupId = voice.id.split('_')[0]!;
+      if (groupId in TTSEngines) {
+        voice.name = name
+          .replace(`${groupId}_`, '')
+          .replace(`${lang}-`, '')
+          .replace('Neural', '')
+          .trim();
+      } else {
+        groupId = 'default';
+      }
+      voice.name = voice.name.replace('NOT_SET', _('Default'));
+      if (!voiceGroups.has(groupId)) {
+        voiceGroups.set(groupId, []);
+      }
+      voiceGroups.get(groupId)!.push(voice);
+    });
+
+    return Array.from(voiceGroups.entries())
+      .map(
+        ([groupId, voices]) =>
+          ({
+            id: groupId,
+            name: TTSEngines[groupId] || groupId,
+            voices: voices,
+            disabled: !this.initialized || voices.length === 0,
+          }) as TTSVoicesGroup,
+      )
+      .sort((a, b) => {
+        if (a.id === 'default') return -1;
+        if (b.id === 'default') return 1;
+        return a.id.localeCompare(b.id);
+      });
+  }
+
+  setPrimaryLang(lang: string) {
+    this.#primaryLang = lang;
+  }
+
+  getGranularities(): TTSGranularity[] {
+    return ['sentence'];
+  }
+
+  getVoiceId() {
+    return this.#currentVoiceId;
+  }
+
+  getSpeakingLang() {
+    return this.#speakingLang;
+  }
+
+  async shutdown() {
+    if (this.#eventListener) {
+      this.#eventListener.unregister();
+      this.#eventListener = null;
+    }
+    await this.stop();
+  }
+}
@@ -1,6 +1,6 @@
-import { TTSGranularity } from '@/types/view';
+import { TTSGranularity, TTSVoice, TTSVoicesGroup } from './types';

-export type TTSMessageCode = 'boundary' | 'error' | 'end';
+type TTSMessageCode = 'boundary' | 'error' | 'end';

 export interface TTSMessageEvent {
  code: TTSMessageCode;
@@ -8,25 +8,21 @@ export interface TTSMessageEvent {
  mark?: string;
 }

-export interface TTSVoice {
-  id: string;
-  name: string;
-  lang: string;
-  disabled?: boolean;
-}
-
 export interface TTSClient {
+  name: string;
+  initialized: boolean;
  init(): Promise<boolean>;
+  shutdown(): Promise<void>;
  speak(ssml: string, signal: AbortSignal, preload?: boolean): AsyncIterable<TTSMessageEvent>;
-  pause(): Promise<void>;
-  resume(): Promise<void>;
+  pause(): Promise<boolean>;
+  resume(): Promise<boolean>;
  stop(): Promise<void>;
  setPrimaryLang(lang: string): void;
  setRate(rate: number): Promise<void>;
  setPitch(pitch: number): Promise<void>;
  setVoice(voice: string): Promise<void>;
  getAllVoices(): Promise<TTSVoice[]>;
-  getVoices(lang: string): Promise<TTSVoice[]>;
+  getVoices(lang: string): Promise<TTSVoicesGroup[]>;
  getGranularities(): TTSGranularity[];
  getVoiceId(): string;
  getSpeakingLang(): string;
@@ -1,50 +1,77 @@
-import { FoliateView, TTSGranularity } from '@/types/view';
-import { TTSClient, TTSMessageCode, TTSVoice } from './TTSClient';
+import { FoliateView } from '@/types/view';
+import { AppService } from '@/types/system';
 import { parseSSMLMarks } from '@/utils/ssml';
+import { TTSGranularity, TTSMark, TTSVoice } from './types';
 import { WebSpeechClient } from './WebSpeechClient';
+import { NativeTTSClient } from './NativeTTSClient';
 import { EdgeTTSClient } from './EdgeTTSClient';
 import { TTSUtils } from './TTSUtils';
-import { TTSMark } from './types';
+import { TTSClient } from './TTSClient';

 type TTSState =
  | 'stopped'
  | 'playing'
  | 'paused'
+  | 'stop-paused'
  | 'backward-paused'
  | 'forward-paused'
  | 'setrate-paused'
  | 'setvoice-paused';

 export class TTSController extends EventTarget {
-  state: TTSState = 'stopped';
+  appService: AppService | null = null;
  view: FoliateView;
  #nossmlCnt: number = 0;
  #currentSpeakAbortController: AbortController | null = null;
  #currentSpeakPromise: Promise<void> | null = null;

+  state: TTSState = 'stopped';
  ttsLang: string = '';
  ttsRate: number = 1.0;
  ttsClient: TTSClient;
  ttsWebClient: TTSClient;
  ttsEdgeClient: TTSClient;
+  ttsNativeClient: TTSClient | null = null;
  ttsWebVoices: TTSVoice[] = [];
  ttsEdgeVoices: TTSVoice[] = [];
+  ttsNativeVoices: TTSVoice[] = [];

-  constructor(view: FoliateView) {
+  constructor(appService: AppService | null, view: FoliateView) {
    super();
    this.ttsWebClient = new WebSpeechClient(this);
    this.ttsEdgeClient = new EdgeTTSClient(this);
+    // TODO: implement native TTS client for iOS and PC
+    if (appService?.isAndroidApp) {
+      this.ttsNativeClient = new NativeTTSClient(this);
+    }
    this.ttsClient = this.ttsWebClient;
+    this.appService = appService;
    this.view = view;
  }

  async init() {
-    await this.ttsWebClient.init();
-    const success = await this.ttsEdgeClient.init();
-    if (success) {
-      this.ttsClient = this.ttsEdgeClient;
-    } else {
-      this.ttsClient = this.ttsWebClient;
+    const availableClients = [];
+    if (this.ttsNativeClient && (await this.ttsNativeClient.init())) {
+      availableClients.push(this.ttsNativeClient);
+      this.ttsNativeVoices = await this.ttsNativeClient.getAllVoices();
+    }
+    if (await this.ttsEdgeClient.init()) {
+      availableClients.push(this.ttsEdgeClient);
+    }
+    if (await this.ttsWebClient.init()) {
+      availableClients.push(this.ttsWebClient);
+    }
+    const preferredClientName = TTSUtils.getPreferredClient();
+    if (preferredClientName) {
+      const preferredClient = availableClients.find(
+        (client) => client.name === preferredClientName,
+      );
+      if (preferredClient) {
+        this.ttsClient = preferredClient;
+      }
+    }
+    if (!this.ttsClient) {
+      this.ttsClient = availableClients[0] || this.ttsWebClient;
    }
    this.ttsWebVoices = await this.ttsWebClient.getAllVoices();
    this.ttsEdgeVoices = await this.ttsEdgeClient.getAllVoices();
@@ -124,7 +151,7 @@ export class TTSController extends EventTarget {
          return await this.forward();
        }
        const iter = await this.ttsClient.speak(ssml, signal);
-        let lastCode: TTSMessageCode = 'boundary';
+        let lastCode;
        for await (const { code, mark } of iter) {
          if (signal.aborted) {
            resolve();
@@ -182,7 +209,10 @@ export class TTSController extends EventTarget {

  async pause() {
    this.state = 'paused';
-    await this.ttsClient.pause().catch((e) => this.error(e));
+    if (!(await this.ttsClient.pause().catch((e) => this.error(e)))) {
+      await this.stop();
+      this.state = 'stop-paused';
+    }
  }

  async resume() {
@@ -197,12 +227,18 @@ export class TTSController extends EventTarget {
    await this.ttsClient.stop().catch((e) => this.error(e));

    if (this.#currentSpeakPromise) {
-      await this.#currentSpeakPromise.catch((e) => this.error(e));
+      const timeout = new Promise((_, reject) =>
+        setTimeout(() => reject(new Error('Stop operation timed out')), 3000),
+      );
+      await Promise.race([this.#currentSpeakPromise.catch((e) => this.error(e)), timeout]).catch(
+        (e) => this.error(e),
+      );
+      this.#currentSpeakPromise = null;
    }
    this.state = 'stopped';
  }

-  // goto previous sentence
+  // goto previous paragraph
  async backward() {
    await this.initViewTTS();
    if (this.state === 'playing') {
@@ -215,7 +251,7 @@ export class TTSController extends EventTarget {
    }
  }

-  // goto next sentence
+  // goto next paragraph
  async forward() {
    await this.initViewTTS();
    if (this.state === 'playing') {
@@ -235,8 +271,9 @@ export class TTSController extends EventTarget {
  }

  async setPrimaryLang(lang: string) {
-    this.ttsEdgeClient.setPrimaryLang(lang);
-    this.ttsWebClient.setPrimaryLang(lang);
+    if (this.ttsEdgeClient.initialized) this.ttsEdgeClient.setPrimaryLang(lang);
+    if (this.ttsWebClient.initialized) this.ttsWebClient.setPrimaryLang(lang);
+    if (this.ttsNativeClient?.initialized) this.ttsNativeClient?.setPrimaryLang(lang);
  }

  async setRate(rate: number) {
@@ -248,7 +285,30 @@ export class TTSController extends EventTarget {
  async getVoices(lang: string) {
    const ttsWebVoices = await this.ttsWebClient.getVoices(lang);
    const ttsEdgeVoices = await this.ttsEdgeClient.getVoices(lang);
-    return [...ttsEdgeVoices, ...ttsWebVoices];
+    const ttsNativeVoices = (await this.ttsNativeClient?.getVoices(lang)) ?? [];
+
+    const voicesGroups = [...ttsNativeVoices, ...ttsEdgeVoices, ...ttsWebVoices];
+    voicesGroups.forEach((group) => {
+      group.voices = group.voices.sort((a, b) => {
+        const aRegion = a.lang.split('-')[1] || '';
+        const bRegion = b.lang.split('-')[1] || '';
+        if (aRegion === bRegion) {
+          return a.name < b.name ? -1 : a.name > b.name ? 1 : 0;
+        }
+        if (aRegion === 'CN') return -1;
+        if (bRegion === 'CN') return 1;
+        if (aRegion === 'TW') return -1;
+        if (bRegion === 'TW') return 1;
+        if (aRegion === 'HK') return -1;
+        if (bRegion === 'HK') return 1;
+        if (aRegion === 'US') return -1;
+        if (bRegion === 'US') return 1;
+        if (aRegion === 'GB') return -1;
+        if (bRegion === 'GB') return 1;
+        return a.name < b.name ? -1 : a.name > b.name ? 1 : 0;
+      });
+    });
+    return voicesGroups;
  }

  async setVoice(voiceId: string, lang: string) {
@@ -256,15 +316,24 @@ export class TTSController extends EventTarget {
    const useEdgeTTS = !!this.ttsEdgeVoices.find(
      (voice) => (voiceId === '' || voice.id === voiceId) && !voice.disabled,
    );
+    const useNativeTTS = !!this.ttsNativeVoices.find(
+      (voice) => (voiceId === '' || voice.id === voiceId) && !voice.disabled,
+    );
    if (useEdgeTTS) {
      this.ttsClient = this.ttsEdgeClient;
      await this.ttsClient.setRate(this.ttsRate);
-      TTSUtils.setPreferredVoice('edge-tts', lang, voiceId);
+    } else if (useNativeTTS) {
+      if (!this.ttsNativeClient) {
+        throw new Error('Native TTS client is not available');
+      }
+      this.ttsClient = this.ttsNativeClient;
+      await this.ttsClient.setRate(this.ttsRate);
    } else {
      this.ttsClient = this.ttsWebClient;
      await this.ttsClient.setRate(this.ttsRate);
-      TTSUtils.setPreferredVoice('web-speech', lang, voiceId);
    }
+    TTSUtils.setPreferredClient(this.ttsClient.name);
+    TTSUtils.setPreferredVoice(this.ttsClient.name, lang, voiceId);
    await this.ttsClient.setVoice(voiceId);
  }

@@ -285,7 +354,16 @@ export class TTSController extends EventTarget {
    this.state = 'stopped';
  }

-  async kill() {
+  async shutdown() {
    await this.stop();
+    if (this.ttsWebClient.initialized) {
+      await this.ttsWebClient.shutdown();
+    }
+    if (this.ttsEdgeClient.initialized) {
+      await this.ttsEdgeClient.shutdown();
+    }
+    if (this.ttsNativeClient?.initialized) {
+      await this.ttsNativeClient.shutdown();
+    }
  }
 }
@@ -1,2 +1,32 @@
 export const SILENCE_DATA =
  'data:audio/mp3;base64,//MkxAAHiAICWABElBeKPL/RANb2w+yiT1g/gTok//lP/W/l3h8QO/OCdCqCW2Cw//MkxAQHkAIWUAhEmAQXWUOFW2dxPu//9mr60ElY5sseQ+xxesmHKtZr7bsqqX2L//MkxAgFwAYiQAhEAC2hq22d3///9FTV6tA36JdgBJoOGgc+7qvqej5Zu7/7uI9l//MkxBQHAAYi8AhEAO193vt9KGOq+6qcT7hhfN5FTInmwk8RkqKImTM55pRQHQSq//MkxBsGkgoIAABHhTACIJLf99nVI///yuW1uBqWfEu7CgNPWGpUadBmZ////4sL//MkxCMHMAH9iABEmAsKioqKigsLCwtVTEFNRTMuOTkuNVVVVVVVVVVVVVVVVVVV//MkxCkECAUYCAAAAFVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVV';
+
+export const WEB_SPEECH_BLACKLISTED_VOICES = [
+  'Albert',
+  'Bad News',
+  'Bahh',
+  'Bells',
+  'Boing',
+  'Bubbles',
+  'Cellos',
+  'Eddy',
+  'Flo',
+  'Fred',
+  'Good News',
+  'Grandma',
+  'Grandpa',
+  'Jester',
+  'Junior',
+  'Kathy',
+  'Organ',
+  'Ralph',
+  'Reed',
+  'Rocko',
+  'Sandy',
+  'Shelley',
+  'Superstar',
+  'Trinoids',
+  'Whisper',
+  'Wobble',
+  'Zarvox',
+];
@@ -1,11 +1,19 @@
 export class TTSUtils {
  private static readonly LOCAL_STORAGE_KEY = 'ttsPreferredVoices';
+  private static readonly PREFERRED_CLIENT_KEY = 'preferredClient';

  private static normalizeLanguage(language: string): string {
    if (!language) return 'n/a';
    return language.toLowerCase().slice(0, 2);
  }

+  static setPreferredClient(engine: string): void {
+    if (!engine) return;
+    const preferences = this.getPreferences();
+    preferences[this.PREFERRED_CLIENT_KEY] = engine;
+    localStorage.setItem(this.LOCAL_STORAGE_KEY, JSON.stringify(preferences));
+  }
+
  static setPreferredVoice(engine: string, language: string, voiceId: string): void {
    if (!engine || !language || !voiceId) return;
    const preferences = this.getPreferences();
@@ -14,6 +22,11 @@ export class TTSUtils {
    localStorage.setItem(this.LOCAL_STORAGE_KEY, JSON.stringify(preferences));
  }

+  static getPreferredClient(): string | null {
+    const preferences = this.getPreferences();
+    return preferences[this.PREFERRED_CLIENT_KEY] || null;
+  }
+
  static getPreferredVoice(engine: string, language: string): string | null {
    const preferences = this.getPreferences();
    const lang = this.normalizeLanguage(language);
@@ -1,40 +1,10 @@
 import { getUserLocale } from '@/utils/misc';
-import { TTSClient, TTSMessageEvent, TTSVoice } from './TTSClient';
+import { TTSClient, TTSMessageEvent } from './TTSClient';
 import { parseSSMLLang, parseSSMLMarks } from '@/utils/ssml';
-import { TTSGranularity } from '@/types/view';
+import { TTSGranularity, TTSMark, TTSVoice, TTSVoicesGroup } from './types';
+import { WEB_SPEECH_BLACKLISTED_VOICES } from './TTSData';
 import { TTSController } from './TTSController';
 import { TTSUtils } from './TTSUtils';
-import { TTSMark } from './types';
-
-const BLACKLISTED_VOICES = [
-  'Albert',
-  'Bad News',
-  'Bahh',
-  'Bells',
-  'Boing',
-  'Bubbles',
-  'Cellos',
-  'Eddy',
-  'Flo',
-  'Fred',
-  'Good News',
-  'Grandma',
-  'Grandpa',
-  'Jester',
-  'Junior',
-  'Kathy',
-  'Organ',
-  'Ralph',
-  'Reed',
-  'Rocko',
-  'Sandy',
-  'Shelley',
-  'Superstar',
-  'Trinoids',
-  'Whisper',
-  'Wobble',
-  'Zarvox',
-];

 interface TTSBoundaryEvent {
  type: 'boundary' | 'end' | 'error';
@@ -52,6 +22,7 @@ async function* speakWithMarks(
  getRate: () => number,
  getPitch: () => number,
  getVoice: (lang: string) => Promise<SpeechSynthesisVoice | null>,
+  setCurrentVoice: (voiceId: string) => void,
  setSpeakingLang: (lang: string) => void,
  dispatchSpeakMark: (mark: TTSMark) => void,
 ) {
@@ -63,7 +34,6 @@ async function* speakWithMarks(
  for (const mark of marks) {
    const { language } = mark;
    const voiceLang = language || defaultLang;
-    setSpeakingLang(voiceLang);
    dispatchSpeakMark(mark);
    utterance.text = mark.text;
    utterance.rate = getRate();
@@ -71,9 +41,11 @@ async function* speakWithMarks(
    const voice = await getVoice(voiceLang);
    if (voice) {
      utterance.voice = voice;
+      setCurrentVoice(voice.voiceURI);
    }
    if (voiceLang) {
      utterance.lang = voiceLang;
+      setSpeakingLang(voiceLang);
    }

    yield {
@@ -102,16 +74,23 @@ async function* speakWithMarks(
  }
 }

+type WebSpeechVoice = SpeechSynthesisVoice & {
+  id: string;
+};
+
 export class WebSpeechClient implements TTSClient {
+  name = 'web-speech';
+  initialized = false;
  controller?: TTSController;
+
+  #voices: WebSpeechVoice[] = [];
  #primaryLang = 'en';
  #speakingLang = '';
+  #currentVoiceId = '';
  #rate = 1.0;
  #pitch = 1.0;
-  #voice: SpeechSynthesisVoice | null = null;
-  #voices: SpeechSynthesisVoice[] = [];
+
  #synth = window.speechSynthesis;
-  available = true;

  constructor(controller?: TTSController) {
    this.controller = controller;
@@ -119,12 +98,16 @@ export class WebSpeechClient implements TTSClient {

  async init() {
    if (!this.#synth) {
-      this.available = false;
-      return this.available;
+      this.initialized = false;
+      return this.initialized;
    }
    await new Promise<void>((resolve) => {
      const populateVoices = () => {
-        this.#voices = this.#synth.getVoices();
+        this.#voices = this.#synth.getVoices().map((voice) => {
+          const webSpeechVoice = voice as WebSpeechVoice;
+          webSpeechVoice.id = voice.voiceURI || voice.name;
+          return webSpeechVoice;
+        });
        // console.log('Voices', this.#voices);
        if (this.#voices.length > 0) {
          resolve();
@@ -140,19 +123,22 @@ export class WebSpeechClient implements TTSClient {
        resolve();
      }
    });
-    return this.available;
+    this.initialized = true;
+    return this.initialized;
  }

-  getVoiceFromLang = async (lang: string) => {
-    const preferredVoiceId = TTSUtils.getPreferredVoice('web-speech', lang);
-    const preferredVoice = this.#voices.find((v) => v.voiceURI === preferredVoiceId);
-    if (preferredVoice) {
-      this.#voice = preferredVoice;
-    } else {
-      const voiceId = (await this.getVoices(lang))[0]?.id ?? '';
-      this.#voice = this.#voices.find((v) => v.voiceURI === voiceId) || null;
-    }
-    return this.#voice;
+  getVoiceIdFromLang = async (lang: string) => {
+    const preferredVoiceId = TTSUtils.getPreferredVoice(this.name, lang);
+    const preferredVoice = this.#voices.find((v) => v.id === preferredVoiceId);
+    const defaultVoice = preferredVoice
+      ? preferredVoice
+      : (await this.getVoices(lang))[0]?.voices[0] || null;
+    return defaultVoice?.id || this.#currentVoiceId || '';
+  };
+
+  getWebSpeechVoiceFromLang = async (lang: string) => {
+    const voiceId = await this.getVoiceIdFromLang(lang);
+    return this.#voices.find((v) => v.id === voiceId) || null;
  };

  async *speak(
@@ -173,7 +159,8 @@ export class WebSpeechClient implements TTSClient {
      defaultLang,
      () => this.#rate,
      () => this.#pitch,
-      this.getVoiceFromLang,
+      this.getWebSpeechVoiceFromLang,
+      (voiceId) => (this.#currentVoiceId = voiceId),
      (lang) => (this.#speakingLang = lang),
      (mark) => this.controller?.dispatchSpeakMark(mark),
    )) {
@@ -198,10 +185,12 @@ export class WebSpeechClient implements TTSClient {

  async pause() {
    this.#synth.pause();
+    return true;
  }

  async resume() {
    this.#synth.resume();
+    return true;
  }

  async stop() {
@@ -223,19 +212,19 @@ export class WebSpeechClient implements TTSClient {
  }

  async setVoice(voiceId: string) {
-    const selectedVoice = this.#voices.find((v) => v.voiceURI === voiceId);
+    const selectedVoice = this.#voices.find((v) => v.id === voiceId);
    if (selectedVoice) {
-      this.#voice = selectedVoice;
+      this.#currentVoiceId = selectedVoice.id;
    }
  }

  async getAllVoices(): Promise<TTSVoice[]> {
    const voices = this.#voices.map((voice) => {
      return {
-        id: voice.voiceURI,
+        id: voice.id,
        name: voice.name,
        lang: voice.lang,
-        disabled: !this.available,
+        disabled: !this.initialized,
      } as TTSVoice;
    });
    return voices;
@@ -247,7 +236,7 @@ export class WebSpeechClient implements TTSClient {
      return !id.includes('com.apple') || id.includes('com.apple.voice.compact');
    };
    const isNotBlacklisted = (voice: SpeechSynthesisVoice) => {
-      return BLACKLISTED_VOICES.some((name) => voice.name.includes(name)) === false;
+      return WEB_SPEECH_BLACKLISTED_VOICES.some((name) => voice.name.includes(name)) === false;
    };
    const filteredVoices = this.#voices
      .filter(
@@ -275,9 +264,16 @@ export class WebSpeechClient implements TTSClient {
        return true;
      });
    voices.forEach((voice) => {
-      voice.disabled = !this.available;
+      voice.disabled = !this.initialized;
    });
-    return voices;
+
+    const voicesGroup: TTSVoicesGroup = {
+      id: 'web-speech-api',
+      name: 'Web TTS',
+      voices,
+      disabled: !this.initialized || voices.length === 0,
+    };
+    return [voicesGroup];
  }

  getGranularities(): TTSGranularity[] {
@@ -287,10 +283,16 @@ export class WebSpeechClient implements TTSClient {
  }

  getVoiceId(): string {
-    return this.#voice?.voiceURI ?? '';
+    return this.#currentVoiceId;
  }

  getSpeakingLang(): string {
    return this.#speakingLang;
  }
+
+  async shutdown(): Promise<void> {
+    this.initialized = false;
+    this.#voices = [];
+    await this.stop();
+  }
 }
@@ -2,5 +2,6 @@ export * from './types';
 export * from './TTSClient';
 export * from './WebSpeechClient';
 export * from './EdgeTTSClient';
+export * from './NativeTTSClient';
 export * from './TTSController';
 export * from './TTSData';
@@ -1,3 +1,19 @@
+export type TTSGranularity = 'sentence' | 'word';
+
+export type TTSVoice = {
+  id: string;
+  name: string;
+  lang: string;
+  disabled?: boolean;
+};
+
+export type TTSVoicesGroup = {
+  id: string;
+  name: string;
+  voices: TTSVoice[];
+  disabled?: boolean;
+};
+
 export type TTSMark = {
  offset: number;
  name: string;
@@ -1,9 +1,8 @@
 import { BookDoc } from '@/libs/document';
 import { BookNote, BookSearchConfig, BookSearchResult } from '@/types/book';
+import { TTSGranularity } from '@/services/tts';
 import { TTS } from 'foliate-js/tts.js';

-export type TTSGranularity = 'sentence' | 'word';
-
 export interface FoliateView extends HTMLElement {
  open: (book: BookDoc) => Promise<void>;
  close: () => void;
@@ -3,6 +3,19 @@ import { TTSMark } from '@/services/tts/types';
 const cleanTextContent = (text: string) =>
  text.replace(/\r\n/g, '  ').replace(/\r/g, ' ').replace(/\n/g, ' ').trimStart();

+export const genSSML = (lang: string, text: string, voice: string, rate: number) => {
+  const cleanedText = text.replace(/^<break\b[^>]*>/i, '');
+  return `
+    <speak version="1.0" xml:lang="${lang}">
+      <voice name="${voice}">
+        <prosody rate="${rate}" >
+            ${cleanedText}
+        </prosody>
+      </voice>
+    </speak>
+  `;
+};
+
 export const parseSSMLMarks = (ssml: string) => {
  const speakMatch = ssml.match(/<speak[^>]*xml:lang="([^"]+)"[^>]*>/i);
  const defaultLang = speakMatch?.[1] ?? 'en';