feat(tts): add auto mode for text-to-speech engine selection

2025-11-02 21:31:13 +05:30
parent da249eaa31
commit cfadeffd24
19 changed files with 579 additions and 154 deletions
--- a/flutter_01.png
+++ b/flutter_01.png
--- a/lib/core/services/settings_service.dart
+++ b/lib/core/services/settings_service.dart
@@ -12,7 +12,7 @@ part 'settings_service.g.dart';
 enum SttPreference { auto, deviceOnly, serverOnly }

 /// TTS engine selection
-enum TtsEngine { device, server }
+enum TtsEngine { auto, device, server }

 /// Service for managing app-wide settings including accessibility preferences
 class SettingsService {
@@ -223,11 +223,15 @@ class SettingsService {

  static TtsEngine _parseTtsEngine(String? raw) {
    switch ((raw ?? '').toLowerCase()) {
+      case 'auto':
+      case '':
+        return TtsEngine.auto;
      case 'server':
        return TtsEngine.server;
      case 'device':
-      default:
        return TtsEngine.device;
+      default:
+        return TtsEngine.auto;
    }
  }

@@ -409,7 +413,7 @@ class AppSettings {
    this.ttsSpeechRate = 0.5,
    this.ttsPitch = 1.0,
    this.ttsVolume = 1.0,
-    this.ttsEngine = TtsEngine.device,
+    this.ttsEngine = TtsEngine.auto,
    this.ttsServerVoiceId,
    this.ttsServerVoiceName,
  });
--- a/lib/features/chat/providers/text_to_speech_provider.dart
+++ b/lib/features/chat/providers/text_to_speech_provider.dart
@@ -107,11 +107,9 @@ class TextToSpeechController extends Notifier<TextToSpeechState> {
    // Listen to settings changes and update TTS when initialized
    ref.listen<AppSettings>(appSettingsProvider, (previous, next) {
      if (_service.isInitialized && _service.isAvailable) {
-        final selectedVoice = next.ttsEngine == TtsEngine.server
-            ? next.ttsServerVoiceId
-            : next.ttsVoice;
        _service.updateSettings(
-          voice: selectedVoice,
+          voice: next.ttsVoice,
+          serverVoice: next.ttsServerVoiceId,
          speechRate: next.ttsSpeechRate,
          pitch: next.ttsPitch,
          volume: next.ttsVolume,
@@ -137,9 +135,8 @@ class TextToSpeechController extends Notifier<TextToSpeechState> {
    final settings = ref.read(appSettingsProvider);
    final future = _service
        .initialize(
-          voice: settings.ttsEngine == TtsEngine.server
-              ? settings.ttsServerVoiceId
-              : settings.ttsVoice,
+          deviceVoice: settings.ttsVoice,
+          serverVoice: settings.ttsServerVoiceId,
          speechRate: settings.ttsSpeechRate,
          pitch: settings.ttsPitch,
          volume: settings.ttsVolume,
--- a/lib/features/chat/services/text_to_speech_service.dart
+++ b/lib/features/chat/services/text_to_speech_service.dart
@@ -16,8 +16,9 @@ class TextToSpeechService {
  final FlutterTts _tts = FlutterTts();
  final AudioPlayer _player = AudioPlayer();
  final ApiService? _api;
-  TtsEngine _engine = TtsEngine.device;
+  TtsEngine _engine = TtsEngine.auto;
  String? _preferredVoice;
+  String? _serverPreferredVoice;
  bool _initialized = false;
  bool _available = false;
  bool _voiceConfigured = false;
@@ -41,6 +42,8 @@ class TextToSpeechService {

  bool get isInitialized => _initialized;
  bool get isAvailable => _available;
+  bool get deviceEngineAvailable => _deviceEngineAvailable;
+  bool get serverEngineAvailable => _api != null;

  TextToSpeechService({ApiService? api}) : _api = api {
    // Wire minimal player events to callbacks
@@ -59,6 +62,69 @@ class TextToSpeechService {
    });
  }

+  Future<void> _configureDeviceEngine({
+    required String? voice,
+    required double speechRate,
+    required double pitch,
+    required double volume,
+  }) async {
+    _deviceEngineAvailable = false;
+    try {
+      await _tts.awaitSpeakCompletion(false);
+      await _tts.setVolume(volume);
+      await _tts.setSpeechRate(speechRate);
+      await _tts.setPitch(pitch);
+
+      if (!kIsWeb && Platform.isIOS) {
+        await _tts.setSharedInstance(true);
+        await _tts.setIosAudioCategory(IosTextToSpeechAudioCategory.playback, [
+          IosTextToSpeechAudioCategoryOptions.mixWithOthers,
+          IosTextToSpeechAudioCategoryOptions.defaultToSpeaker,
+          IosTextToSpeechAudioCategoryOptions.allowBluetooth,
+          IosTextToSpeechAudioCategoryOptions.allowBluetoothA2DP,
+        ]);
+      }
+
+      if (_engine != TtsEngine.server) {
+        await _setVoiceByName(_preferredVoice);
+      } else {
+        _voiceConfigured = false;
+      }
+
+      _deviceEngineAvailable = true;
+    } catch (e) {
+      _voiceConfigured = false;
+      _deviceEngineAvailable = false;
+      rethrow;
+    }
+  }
+
+  bool _computeAvailability() {
+    final serverAvailable = _api != null;
+    switch (_engine) {
+      case TtsEngine.device:
+        return _deviceEngineAvailable;
+      case TtsEngine.server:
+        return serverAvailable;
+      case TtsEngine.auto:
+        return _deviceEngineAvailable || serverAvailable;
+    }
+  }
+
+  bool _shouldUseServer() {
+    if (_engine == TtsEngine.server) {
+      return _api != null;
+    }
+    if (_engine == TtsEngine.device) {
+      return false;
+    }
+    // Auto: prefer device when available, otherwise fall back to server
+    if (_deviceEngineAvailable) {
+      return false;
+    }
+    return _api != null;
+  }
+
  /// Register callbacks for TTS lifecycle events
  void bindHandlers({
    VoidCallback? onStart,
@@ -96,56 +162,58 @@ class TextToSpeechService {

  /// Initialize the native TTS engine lazily
  Future<bool> initialize({
-    String? voice,
+    String? deviceVoice,
+    String? serverVoice,
    double speechRate = 0.5,
    double pitch = 1.0,
    double volume = 1.0,
-    TtsEngine engine = TtsEngine.device,
+    TtsEngine engine = TtsEngine.auto,
  }) async {
    if (_initialized) {
+      _engine = engine;
+      if (deviceVoice != null) {
+        _preferredVoice = deviceVoice;
+        _voiceConfigured = false;
+      }
+      if (serverVoice != null) {
+        _serverPreferredVoice = serverVoice;
+      }
+      _available = _computeAvailability();
      return _available;
    }

-    try {
    _engine = engine;
-      _preferredVoice = voice;
-      await _tts.awaitSpeakCompletion(false);
+    _preferredVoice = deviceVoice;
+    _serverPreferredVoice = serverVoice;
+    _voiceConfigured = false;

-      // Set volume
-      await _tts.setVolume(volume);
-
-      // Set speech rate
-      await _tts.setSpeechRate(speechRate);
-
-      // Set pitch
-      await _tts.setPitch(pitch);
-
-      if (!kIsWeb && Platform.isIOS) {
-        await _tts.setSharedInstance(true);
-        await _tts.setIosAudioCategory(IosTextToSpeechAudioCategory.playback, [
-          IosTextToSpeechAudioCategoryOptions.mixWithOthers,
-          IosTextToSpeechAudioCategoryOptions.defaultToSpeaker,
-          IosTextToSpeechAudioCategoryOptions.allowBluetooth,
-          IosTextToSpeechAudioCategoryOptions.allowBluetoothA2DP,
-        ]);
-      }
-
-      // Set the voice (specific or default) when using device engine
-      if (_engine == TtsEngine.device) {
-        await _setVoiceByName(voice);
-      }
-      _deviceEngineAvailable = true;
+    if (_engine != TtsEngine.server || _api == null) {
+      try {
+        await _configureDeviceEngine(
+          voice: deviceVoice,
+          speechRate: speechRate,
+          pitch: pitch,
+          volume: volume,
+        );
      } catch (e) {
-      _deviceEngineAvailable = false;
-      if (_engine != TtsEngine.server) {
+        if (_engine == TtsEngine.device) {
          _available = false;
          _onError?.call(e.toString());
          _initialized = true;
          return _available;
        }
      }
+    } else {
+      _deviceEngineAvailable = false;
+      try {
+        await _tts.awaitSpeakCompletion(false);
+        await _tts.setVolume(volume);
+        await _tts.setSpeechRate(speechRate);
+        await _tts.setPitch(pitch);
+      } catch (_) {}
+    }

-    _available = _engine == TtsEngine.server || _deviceEngineAvailable;
+    _available = _computeAvailability();
    _initialized = true;
    return _available;
  }
@@ -156,10 +224,23 @@ class TextToSpeechService {
    }

    if (!_initialized) {
-      await initialize(voice: _preferredVoice, engine: _engine);
+      await initialize(
+        deviceVoice: _preferredVoice,
+        serverVoice: _serverPreferredVoice,
+        engine: _engine,
+      );
    }

-    if (_engine == TtsEngine.server && _api != null) {
+    final bool useServer = _shouldUseServer();
+
+    if (useServer) {
+      if (_api == null) {
+        if (_deviceEngineAvailable) {
+          await _speakOnDevice(text);
+          return;
+        }
+        throw StateError('Server text-to-speech is unavailable');
+      }
      // Server-backed TTS with sentence chunking & queued playback
      try {
        await _startServerChunkedPlayback(text);
@@ -196,7 +277,7 @@ class TextToSpeechService {
  Future<void> pause() async {
    if (!_initialized) return;
    try {
-      if (_engine == TtsEngine.server) {
+      if (_shouldUseServer()) {
        await _player.pause();
        _handlePause();
      } else if (_deviceEngineAvailable) {
@@ -210,7 +291,7 @@ class TextToSpeechService {
  Future<void> resume() async {
    if (!_initialized) return;
    try {
-      if (_engine == TtsEngine.server) {
+      if (_shouldUseServer()) {
        if (_waitingNext && (_currentIndex + 1) < _buffered.length) {
          _waitingNext = false;
          await _playNextIfBuffered(_session);
@@ -235,7 +316,7 @@ class TextToSpeechService {
      _expectedChunks = 0;
      _currentIndex = -1;
      _waitingNext = false;
-      if (_engine == TtsEngine.server) {
+      if (_shouldUseServer()) {
        await _player.stop();
        _handleCancel();
      } else {
@@ -254,17 +335,23 @@ class TextToSpeechService {
  /// Update TTS settings on-the-fly
  Future<void> updateSettings({
    Object? voice = const _VoiceNotProvided(),
+    Object? serverVoice = const _VoiceNotProvided(),
    double? speechRate,
    double? pitch,
    double? volume,
    TtsEngine? engine,
  }) async {
    final voiceProvided = voice is! _VoiceNotProvided;
+    final serverVoiceProvided = serverVoice is! _VoiceNotProvided;
    final voiceValue = voiceProvided ? voice as String? : null;
+    final serverVoiceValue = serverVoiceProvided
+        ? serverVoice as String?
+        : null;
    if (!_initialized || !_available) {
      // Allow engine and voice to update before init
      if (engine != null) _engine = engine;
      if (voiceProvided) _preferredVoice = voiceValue;
+      if (serverVoiceProvided) _serverPreferredVoice = serverVoiceValue;
      return;
    }

@@ -275,6 +362,9 @@ class TextToSpeechService {
      if (voiceProvided) {
        _preferredVoice = voiceValue;
      }
+      if (serverVoiceProvided) {
+        _serverPreferredVoice = serverVoiceValue;
+      }
      if (volume != null) {
        await _tts.setVolume(volume);
      }
@@ -284,13 +374,15 @@ class TextToSpeechService {
      if (pitch != null) {
        await _tts.setPitch(pitch);
      }
-      // Set specific voice by name on device engine
-      if (_engine == TtsEngine.device && voiceProvided) {
+      // Set specific voice by name on device-capable engines
+      if (_engine != TtsEngine.server && voiceProvided) {
        await _setVoiceByName(_preferredVoice);
      }
    } catch (e) {
      _onError?.call(e.toString());
    }
+
+    _available = _computeAvailability();
  }

  /// Set voice by name, or use system default if null
@@ -343,7 +435,11 @@ class TextToSpeechService {
  /// Get available voices from the TTS engine
  Future<List<Map<String, dynamic>>> getAvailableVoices() async {
    if (!_initialized) {
-      await initialize(voice: _preferredVoice, engine: _engine);
+      await initialize(
+        deviceVoice: _preferredVoice,
+        serverVoice: _serverPreferredVoice,
+        engine: _engine,
+      );
    }

    if (_engine == TtsEngine.server && _api != null) {
@@ -425,6 +521,10 @@ class TextToSpeechService {
  }

  Future<String?> _resolveServerVoice() async {
+    final serverSelected = _serverPreferredVoice?.trim();
+    if (serverSelected != null && serverSelected.isNotEmpty) {
+      return serverSelected;
+    }
    final selected = _preferredVoice?.trim();
    if (selected != null && selected.isNotEmpty) {
      return selected;
--- a/lib/features/chat/services/voice_call_service.dart
+++ b/lib/features/chat/services/voice_call_service.dart
@@ -132,9 +132,8 @@ class VoiceCallService {
    // Initialize TTS with current app settings (engine/voice/rate/pitch/volume)
    final settings = _ref.read(appSettingsProvider);
    await _tts.initialize(
-      voice: settings.ttsEngine == TtsEngine.server
-          ? settings.ttsServerVoiceId
-          : settings.ttsVoice,
+      deviceVoice: settings.ttsVoice,
+      serverVoice: settings.ttsServerVoiceId,
      speechRate: settings.ttsSpeechRate,
      pitch: settings.ttsPitch,
      volume: settings.ttsVolume,
@@ -587,11 +586,9 @@ VoiceCallService voiceCallService(Ref ref) {
  // Keep TTS settings in sync with app settings during a call
  ref.listen<AppSettings>(appSettingsProvider, (previous, next) {
    // Update voice/engine and runtime parameters
-    final selectedVoice = next.ttsEngine == TtsEngine.server
-        ? next.ttsServerVoiceId
-        : next.ttsVoice;
    service._tts.updateSettings(
-      voice: selectedVoice,
+      voice: next.ttsVoice,
+      serverVoice: next.ttsServerVoiceId,
      speechRate: next.ttsSpeechRate,
      pitch: next.ttsPitch,
      volume: next.ttsVolume,
--- a/lib/features/profile/views/app_customization_page.dart
+++ b/lib/features/profile/views/app_customization_page.dart
@@ -698,6 +698,35 @@ class AppCustomizationPage extends ConsumerWidget {
  ) {
    final theme = context.conduitTheme;
    final l10n = AppLocalizations.of(context)!;
+    final ttsService = ref.watch(textToSpeechServiceProvider);
+    final bool deviceAvailable =
+        ttsService.deviceEngineAvailable || !ttsService.isInitialized;
+    final bool serverAvailable = ttsService.serverEngineAvailable;
+    final bool autoSelectable = deviceAvailable || serverAvailable;
+    final bool deviceSelectable = deviceAvailable;
+    final bool serverSelectable = serverAvailable;
+    final ttsDescription = _ttsPreferenceDescription(l10n, settings);
+    final warnings = <String>[];
+    switch (settings.ttsEngine) {
+      case TtsEngine.auto:
+        if (!deviceAvailable) {
+          warnings.add(l10n.ttsDeviceUnavailableWarning);
+        }
+        if (!serverAvailable) {
+          warnings.add(l10n.ttsServerUnavailableWarning);
+        }
+        break;
+      case TtsEngine.device:
+        if (!deviceAvailable) {
+          warnings.add(l10n.ttsDeviceUnavailableWarning);
+        }
+        break;
+      case TtsEngine.server:
+        if (!serverAvailable) {
+          warnings.add(l10n.ttsServerUnavailableWarning);
+        }
+        break;
+    }
    return Column(
      crossAxisAlignment: CrossAxisAlignment.start,
      children: [
@@ -733,10 +762,44 @@ class AppCustomizationPage extends ConsumerWidget {
                        ) ??
                        TextStyle(color: theme.sidebarForeground, fontSize: 14),
                  ),
-                  const Spacer(),
+                ],
+              ),
+              const SizedBox(height: Spacing.sm),
              Wrap(
                spacing: Spacing.sm,
+                runSpacing: Spacing.sm,
                children: [
+                  ChoiceChip(
+                    label: Text(l10n.ttsEngineAuto),
+                    selected: settings.ttsEngine == TtsEngine.auto,
+                    showCheckmark: false,
+                    selectedColor: theme.buttonPrimary,
+                    backgroundColor: theme.cardBackground,
+                    side: BorderSide(
+                      color: settings.ttsEngine == TtsEngine.auto
+                          ? theme.buttonPrimary.withValues(alpha: 0.6)
+                          : theme.textPrimary.withValues(
+                              alpha: autoSelectable ? 0.2 : 0.12,
+                            ),
+                    ),
+                    labelStyle: TextStyle(
+                      color: settings.ttsEngine == TtsEngine.auto
+                          ? theme.buttonPrimaryText
+                          : theme.textPrimary.withValues(
+                              alpha: autoSelectable ? 1.0 : 0.45,
+                            ),
+                      fontWeight: FontWeight.w600,
+                    ),
+                    onSelected: autoSelectable
+                        ? (value) {
+                            if (value) {
+                              ref
+                                  .read(appSettingsProvider.notifier)
+                                  .setTtsEngine(TtsEngine.auto);
+                            }
+                          }
+                        : null,
+                  ),
                  ChoiceChip(
                    label: Text(l10n.ttsEngineDevice),
                    selected: settings.ttsEngine == TtsEngine.device,
@@ -746,23 +809,27 @@ class AppCustomizationPage extends ConsumerWidget {
                    side: BorderSide(
                      color: settings.ttsEngine == TtsEngine.device
                          ? theme.buttonPrimary.withValues(alpha: 0.6)
-                              : theme.textPrimary.withValues(alpha: 0.2),
+                          : theme.textPrimary.withValues(
+                              alpha: deviceSelectable ? 0.2 : 0.12,
+                            ),
                    ),
                    labelStyle: TextStyle(
                      color: settings.ttsEngine == TtsEngine.device
                          ? theme.buttonPrimaryText
-                              : theme.textPrimary,
+                          : theme.textPrimary.withValues(
+                              alpha: deviceSelectable ? 1.0 : 0.45,
+                            ),
                      fontWeight: FontWeight.w600,
                    ),
-                        onSelected: (v) {
-                          if (v) {
-                            final notifier = ref.read(
-                              appSettingsProvider.notifier,
-                            );
-                            notifier.setTtsEngine(TtsEngine.device);
-                            // Keep previous voice (device voices)
+                    onSelected: deviceSelectable
+                        ? (value) {
+                            if (value) {
+                              ref
+                                  .read(appSettingsProvider.notifier)
+                                  .setTtsEngine(TtsEngine.device);
                            }
-                        },
+                          }
+                        : null,
                  ),
                  ChoiceChip(
                    label: Text(l10n.ttsEngineServer),
@@ -773,42 +840,82 @@ class AppCustomizationPage extends ConsumerWidget {
                    side: BorderSide(
                      color: settings.ttsEngine == TtsEngine.server
                          ? theme.buttonPrimary.withValues(alpha: 0.6)
-                              : theme.textPrimary.withValues(alpha: 0.2),
+                          : theme.textPrimary.withValues(
+                              alpha: serverSelectable ? 0.2 : 0.12,
+                            ),
                    ),
                    labelStyle: TextStyle(
                      color: settings.ttsEngine == TtsEngine.server
                          ? theme.buttonPrimaryText
-                              : theme.textPrimary,
+                          : theme.textPrimary.withValues(
+                              alpha: serverSelectable ? 1.0 : 0.45,
+                            ),
                      fontWeight: FontWeight.w600,
                    ),
-                        onSelected: (v) {
-                          if (v) {
+                    onSelected: serverSelectable
+                        ? (value) {
+                            if (value) {
                              final notifier = ref.read(
                                appSettingsProvider.notifier,
                              );
-                            // Clear device-specific voice so server can default
                              notifier.setTtsVoice(null);
                              notifier.setTtsEngine(TtsEngine.server);
                            }
-                        },
+                          }
+                        : null,
                  ),
                ],
              ),
-                ],
+              const SizedBox(height: Spacing.sm),
+              AnimatedSwitcher(
+                duration: const Duration(milliseconds: 200),
+                child: Text(
+                  ttsDescription,
+                  key: ValueKey<String>(
+                    'tts-desc-${settings.ttsEngine.name}',
                  ),
+                  style:
+                      theme.bodyMedium?.copyWith(
+                        color: theme.sidebarForeground.withValues(
+                          alpha: 0.9,
+                        ),
+                      ) ??
+                      TextStyle(
+                        color: theme.sidebarForeground.withValues(
+                          alpha: 0.9,
+                        ),
+                        fontSize: 14,
+                      ),
+                ),
+              ),
+              if (warnings.isNotEmpty) ...[
+                const SizedBox(height: Spacing.sm),
+                ...warnings.map(
+                  (warning) => Padding(
+                    padding: const EdgeInsets.only(top: Spacing.xs),
+                    child: Text(
+                      warning,
+                      style:
+                          theme.bodySmall?.copyWith(
+                            color: theme.error,
+                            fontWeight: FontWeight.w600,
+                          ) ??
+                          TextStyle(
+                            color: theme.error,
+                            fontSize: 12,
+                            fontWeight: FontWeight.w600,
+                          ),
+                    ),
+                  ),
+                ),
+              ],
            ],
          ),
        ),
        const SizedBox(height: Spacing.sm),
        _ExpandableCard(
          title: l10n.ttsVoice,
-          subtitle: _getDisplayVoiceName(
-            settings.ttsEngine == TtsEngine.server
-                ? ((settings.ttsServerVoiceName ?? settings.ttsServerVoiceId) ??
-                      '')
-                : (settings.ttsVoice ?? ''),
-            l10n.ttsSystemDefault,
-          ),
+          subtitle: _ttsVoiceSubtitle(l10n, settings),
          icon: UiUtils.platformIcon(
            ios: CupertinoIcons.speaker_3,
            android: Icons.record_voice_over,
@@ -827,14 +934,7 @@ class AppCustomizationPage extends ConsumerWidget {
                  color: theme.buttonPrimary,
                ),
                title: l10n.ttsVoice,
-                subtitle: _getDisplayVoiceName(
-                  settings.ttsEngine == TtsEngine.server
-                      ? ((settings.ttsServerVoiceName ??
-                                settings.ttsServerVoiceId) ??
-                            '')
-                      : (settings.ttsVoice ?? ''),
-                  l10n.ttsSystemDefault,
-                ),
+                subtitle: _ttsVoiceSubtitle(l10n, settings),
                onTap: () => _showVoicePickerSheet(context, ref, settings),
              ),
              const SizedBox(height: Spacing.md),
@@ -928,6 +1028,39 @@ class AppCustomizationPage extends ConsumerWidget {
    }
  }

+  String _ttsPreferenceDescription(
+    AppLocalizations l10n,
+    AppSettings settings,
+  ) {
+    switch (settings.ttsEngine) {
+      case TtsEngine.auto:
+        return l10n.ttsEngineAutoDescription;
+      case TtsEngine.device:
+        return l10n.ttsEngineDeviceDescription;
+      case TtsEngine.server:
+        return l10n.ttsEngineServerDescription;
+    }
+  }
+
+  String _ttsVoiceSubtitle(AppLocalizations l10n, AppSettings settings) {
+    final deviceName = _getDisplayVoiceName(
+      settings.ttsVoice,
+      l10n.ttsSystemDefault,
+    );
+    final serverVoice =
+        (settings.ttsServerVoiceName ?? settings.ttsServerVoiceId) ?? '';
+    final serverName = _getDisplayVoiceName(serverVoice, l10n.ttsSystemDefault);
+
+    switch (settings.ttsEngine) {
+      case TtsEngine.auto:
+        return '${l10n.ttsEngineDevice}: $deviceName • ${l10n.ttsEngineServer}: $serverName';
+      case TtsEngine.device:
+        return deviceName;
+      case TtsEngine.server:
+        return serverName;
+    }
+  }
+
  Widget _buildSliderTile(
    BuildContext context,
    WidgetRef ref, {
--- a/lib/l10n/app_de.arb
+++ b/lib/l10n/app_de.arb
@@ -314,10 +314,16 @@
  "sttEngineServer": "Server",
  "sttEngineAutoDescription": "Verwendet die Erkennung auf dem Gerät, wenn verfügbar, und greift sonst auf deinen Server zurück.",
  "sttEngineDeviceDescription": "Behält Audio auf diesem Gerät. Spracheingabe funktioniert nicht, wenn das Gerät keine Spracherkennung unterstützt.",
-  "sttEngineServerDescription": "Sendet Aufnahmen immer an deinen Conduit-Server zur Transkription.",
+  "sttEngineServerDescription": "Sendet Aufnahmen immer an deinen OpenWebUI-Server zur Transkription.",
  "sttDeviceUnavailableWarning": "Auf diesem Gerät steht keine Spracherkennung zur Verfügung.",
  "sttServerUnavailableWarning": "Verbinde dich mit einem Server mit aktivierter Transkription, um diese Option zu nutzen.",
  "ttsSettings": "Text zu Sprache",
+  "ttsEngineAuto": "Automatisch",
+  "ttsEngineAutoDescription": "Verwendet die Sprachausgabe auf dem Gerät, wenn verfügbar, und greift sonst auf deinen Server zurück.",
+  "ttsEngineDeviceDescription": "Behält die Ausgabe auf diesem Gerät. Sprachausgabe funktioniert nicht, wenn das Gerät keine TTS-Unterstützung bietet.",
+  "ttsEngineServerDescription": "Sendet die Ausgabe immer an deinen OpenWebUI-Server.",
+  "ttsDeviceUnavailableWarning": "Sprachausgabe auf dem Gerät steht auf diesem Gerät nicht zur Verfügung.",
+  "ttsServerUnavailableWarning": "Verbinde dich mit einem Server mit aktivierter Sprachausgabe, um diese Option zu nutzen.",
  "ttsVoice": "Stimme",
  "ttsSpeechRate": "Sprechgeschwindigkeit",
  "ttsPitch": "Tonhöhe",
--- a/lib/l10n/app_en.arb
+++ b/lib/l10n/app_en.arb
@@ -1247,7 +1247,7 @@
  "@sttEngineDeviceDescription": {
    "description": "Description shown when on-device speech-to-text preference is active."
  },
-  "sttEngineServerDescription": "Always send recordings to your Conduit server for transcription.",
+  "sttEngineServerDescription": "Always send recordings to your OpenWebUI server for transcription.",
  "@sttEngineServerDescription": {
    "description": "Description shown when server speech-to-text preference is active."
  },
@@ -1263,6 +1263,10 @@
  "@ttsEngineLabel": {
    "description": "Label for selecting the text-to-speech engine."
  },
+  "ttsEngineAuto": "Auto",
+  "@ttsEngineAuto": {
+    "description": "Chip label for automatically selecting the text-to-speech engine."
+  },
  "ttsEngineDevice": "On device",
  "@ttsEngineDevice": {
    "description": "Chip label for using on-device text-to-speech."
@@ -1271,6 +1275,26 @@
  "@ttsEngineServer": {
    "description": "Chip label for using server-side text-to-speech."
  },
+  "ttsEngineAutoDescription": "Use on-device speech when available and fall back to your server.",
+  "@ttsEngineAutoDescription": {
+    "description": "Description shown when automatic text-to-speech preference is active."
+  },
+  "ttsEngineDeviceDescription": "Keep synthesis on this device. Voice playback stops working if on-device TTS isn’t supported.",
+  "@ttsEngineDeviceDescription": {
+    "description": "Description shown when on-device text-to-speech preference is active."
+  },
+  "ttsEngineServerDescription": "Always request audio from your OpenWebUI server.",
+  "@ttsEngineServerDescription": {
+    "description": "Description shown when server text-to-speech preference is active."
+  },
+  "ttsDeviceUnavailableWarning": "On-device text-to-speech isn’t available on this device.",
+  "@ttsDeviceUnavailableWarning": {
+    "description": "Warning shown when on-device text-to-speech is unavailable."
+  },
+  "ttsServerUnavailableWarning": "Connect to a server with text-to-speech enabled to use this option.",
+  "@ttsServerUnavailableWarning": {
+    "description": "Warning shown when server text-to-speech is unavailable."
+  },
  "ttsSettings": "Text to Speech",
  "@ttsSettings": {
    "description": "Section header for TTS-related customization options."
--- a/lib/l10n/app_es.arb
+++ b/lib/l10n/app_es.arb
@@ -314,10 +314,16 @@
  "sttEngineServer": "Servidor",
  "sttEngineAutoDescription": "Usa el reconocimiento en el dispositivo cuando esté disponible y, si no, recurre a tu servidor.",
  "sttEngineDeviceDescription": "Mantiene el audio en este dispositivo. La entrada de voz no funciona si el dispositivo no admite reconocimiento de voz.",
-  "sttEngineServerDescription": "Envía siempre las grabaciones a tu servidor Conduit para la transcripción.",
+  "sttEngineServerDescription": "Envía siempre las grabaciones a tu servidor OpenWebUI para la transcripción.",
  "sttDeviceUnavailableWarning": "El reconocimiento de voz en el dispositivo no está disponible en este dispositivo.",
  "sttServerUnavailableWarning": "Conéctate a un servidor con transcripción habilitada para usar esta opción.",
  "ttsSettings": "Texto a voz",
+  "ttsEngineAuto": "Automático",
+  "ttsEngineAutoDescription": "Usa la síntesis en el dispositivo cuando esté disponible y, si no, recurre a tu servidor.",
+  "ttsEngineDeviceDescription": "Mantiene la síntesis en este dispositivo. La reproducción de voz no funciona si el dispositivo no admite TTS.",
+  "ttsEngineServerDescription": "Solicita siempre el audio a tu servidor OpenWebUI.",
+  "ttsDeviceUnavailableWarning": "La síntesis de voz en el dispositivo no está disponible en este dispositivo.",
+  "ttsServerUnavailableWarning": "Conéctate a un servidor con texto a voz habilitado para usar esta opción.",
  "ttsVoice": "Voz",
  "ttsSpeechRate": "Velocidad de voz",
  "ttsPitch": "Tono",
--- a/lib/l10n/app_fr.arb
+++ b/lib/l10n/app_fr.arb
@@ -314,10 +314,16 @@
  "sttEngineServer": "Serveur",
  "sttEngineAutoDescription": "Utilise la reconnaissance sur l’appareil quand c’est possible, sinon bascule vers votre serveur.",
  "sttEngineDeviceDescription": "Conserve l’audio sur cet appareil. L’entrée vocale cesse de fonctionner si la reconnaissance vocale n’est pas prise en charge.",
-  "sttEngineServerDescription": "Envoie toujours les enregistrements à votre serveur Conduit pour transcription.",
+  "sttEngineServerDescription": "Envoie toujours les enregistrements à votre serveur OpenWebUI pour transcription.",
  "sttDeviceUnavailableWarning": "La reconnaissance vocale sur l’appareil n’est pas disponible sur cet appareil.",
  "sttServerUnavailableWarning": "Connectez-vous à un serveur avec la transcription activée pour utiliser cette option.",
-  "ttsSettings": "Synthèse vocale",
+"ttsSettings": "Synthèse vocale",
+  "ttsEngineAuto": "Auto",
+  "ttsEngineAutoDescription": "Utilise la synthèse locale quand c’est possible, sinon bascule vers votre serveur.",
+  "ttsEngineDeviceDescription": "Garde la synthèse sur cet appareil. La lecture vocale ne fonctionne plus si l’appareil n’offre pas la synthèse vocale.",
+  "ttsEngineServerDescription": "Demande toujours l'audio à votre serveur OpenWebUI.",
+  "ttsDeviceUnavailableWarning": "La synthèse vocale sur l’appareil n’est pas disponible sur cet appareil.",
+  "ttsServerUnavailableWarning": "Connectez-vous à un serveur avec la synthèse vocale activée pour utiliser cette option.",
  "ttsVoice": "Voix",
  "ttsSpeechRate": "Vitesse de parole",
  "ttsPitch": "Hauteur",
--- a/lib/l10n/app_it.arb
+++ b/lib/l10n/app_it.arb
@@ -314,10 +314,16 @@
  "sttEngineServer": "Server",
  "sttEngineAutoDescription": "Usa il riconoscimento sul dispositivo quando disponibile e altrimenti passa al tuo server.",
  "sttEngineDeviceDescription": "Mantiene l’audio su questo dispositivo. L’input vocale non funziona se il dispositivo non supporta il riconoscimento vocale.",
-  "sttEngineServerDescription": "Invia sempre le registrazioni al tuo server Conduit per la trascrizione.",
+  "sttEngineServerDescription": "Invia sempre le registrazioni al tuo server OpenWebUI per la trascrizione.",
  "sttDeviceUnavailableWarning": "Il riconoscimento vocale sul dispositivo non è disponibile su questo dispositivo.",
  "sttServerUnavailableWarning": "Collegati a un server con la trascrizione abilitata per usare questa opzione.",
-  "ttsSettings": "Sintesi vocale",
+"ttsSettings": "Sintesi vocale",
+  "ttsEngineAuto": "Automatico",
+  "ttsEngineAutoDescription": "Usa la sintesi sul dispositivo quando disponibile e altrimenti passa al tuo server.",
+  "ttsEngineDeviceDescription": "Mantiene la sintesi su questo dispositivo. La riproduzione vocale non funziona se il dispositivo non supporta il TTS.",
+  "ttsEngineServerDescription": "Richiede sempre l'audio dal tuo server OpenWebUI.",
+  "ttsDeviceUnavailableWarning": "La sintesi vocale sul dispositivo non è disponibile su questo dispositivo.",
+  "ttsServerUnavailableWarning": "Collegati a un server con la sintesi vocale abilitata per usare questa opzione.",
  "ttsVoice": "Voce",
  "ttsSpeechRate": "Velocità di sintesi vocale",
  "ttsPitch": "Tonalità",
--- a/lib/l10n/app_localizations.dart
+++ b/lib/l10n/app_localizations.dart
@@ -1829,7 +1829,7 @@ abstract class AppLocalizations {
  /// Description shown when server speech-to-text preference is active.
  ///
  /// In en, this message translates to:
-  /// **'Always send recordings to your Conduit server for transcription.'**
+  /// **'Always send recordings to your OpenWebUI server for transcription.'**
  String get sttEngineServerDescription;

  /// Warning shown when the user selects on-device speech recognition but it is unavailable.
@@ -1850,6 +1850,12 @@ abstract class AppLocalizations {
  /// **'Engine'**
  String get ttsEngineLabel;

+  /// Chip label for automatically selecting the text-to-speech engine.
+  ///
+  /// In en, this message translates to:
+  /// **'Auto'**
+  String get ttsEngineAuto;
+
  /// Chip label for using on-device text-to-speech.
  ///
  /// In en, this message translates to:
@@ -1862,6 +1868,36 @@ abstract class AppLocalizations {
  /// **'Server'**
  String get ttsEngineServer;

+  /// Description shown when automatic text-to-speech preference is active.
+  ///
+  /// In en, this message translates to:
+  /// **'Use on-device speech when available and fall back to your server.'**
+  String get ttsEngineAutoDescription;
+
+  /// Description shown when on-device text-to-speech preference is active.
+  ///
+  /// In en, this message translates to:
+  /// **'Keep synthesis on this device. Voice playback stops working if on-device TTS isn’t supported.'**
+  String get ttsEngineDeviceDescription;
+
+  /// Description shown when server text-to-speech preference is active.
+  ///
+  /// In en, this message translates to:
+  /// **'Always request audio from your OpenWebUI server.'**
+  String get ttsEngineServerDescription;
+
+  /// Warning shown when on-device text-to-speech is unavailable.
+  ///
+  /// In en, this message translates to:
+  /// **'On-device text-to-speech isn’t available on this device.'**
+  String get ttsDeviceUnavailableWarning;
+
+  /// Warning shown when server text-to-speech is unavailable.
+  ///
+  /// In en, this message translates to:
+  /// **'Connect to a server with text-to-speech enabled to use this option.'**
+  String get ttsServerUnavailableWarning;
+
  /// Section header for TTS-related customization options.
  ///
  /// In en, this message translates to:
--- a/lib/l10n/app_localizations_de.dart
+++ b/lib/l10n/app_localizations_de.dart
@@ -956,7 +956,7 @@ class AppLocalizationsDe extends AppLocalizations {

  @override
  String get sttEngineServerDescription =>
-      'Sendet Aufnahmen immer an deinen Conduit-Server zur Transkription.';
+      'Sendet Aufnahmen immer an deinen OpenWebUI-Server zur Transkription.';

  @override
  String get sttDeviceUnavailableWarning =>
@@ -969,12 +969,35 @@ class AppLocalizationsDe extends AppLocalizations {
  @override
  String get ttsEngineLabel => 'Engine';

+  @override
+  String get ttsEngineAuto => 'Automatisch';
+
  @override
  String get ttsEngineDevice => 'Auf dem Gerät';

  @override
  String get ttsEngineServer => 'Server';

+  @override
+  String get ttsEngineAutoDescription =>
+      'Verwendet die Sprachausgabe auf dem Gerät, wenn verfügbar, und greift sonst auf deinen Server zurück.';
+
+  @override
+  String get ttsEngineDeviceDescription =>
+      'Behält die Ausgabe auf diesem Gerät. Sprachausgabe funktioniert nicht, wenn das Gerät keine TTS-Unterstützung bietet.';
+
+  @override
+  String get ttsEngineServerDescription =>
+      'Sendet die Ausgabe immer an deinen OpenWebUI-Server.';
+
+  @override
+  String get ttsDeviceUnavailableWarning =>
+      'Sprachausgabe auf dem Gerät steht auf diesem Gerät nicht zur Verfügung.';
+
+  @override
+  String get ttsServerUnavailableWarning =>
+      'Verbinde dich mit einem Server mit aktivierter Sprachausgabe, um diese Option zu nutzen.';
+
  @override
  String get ttsSettings => 'Text zu Sprache';

--- a/lib/l10n/app_localizations_en.dart
+++ b/lib/l10n/app_localizations_en.dart
@@ -950,7 +950,7 @@ class AppLocalizationsEn extends AppLocalizations {

  @override
  String get sttEngineServerDescription =>
-      'Always send recordings to your Conduit server for transcription.';
+      'Always send recordings to your OpenWebUI server for transcription.';

  @override
  String get sttDeviceUnavailableWarning =>
@@ -963,12 +963,35 @@ class AppLocalizationsEn extends AppLocalizations {
  @override
  String get ttsEngineLabel => 'Engine';

+  @override
+  String get ttsEngineAuto => 'Auto';
+
  @override
  String get ttsEngineDevice => 'On device';

  @override
  String get ttsEngineServer => 'Server';

+  @override
+  String get ttsEngineAutoDescription =>
+      'Use on-device speech when available and fall back to your server.';
+
+  @override
+  String get ttsEngineDeviceDescription =>
+      'Keep synthesis on this device. Voice playback stops working if on-device TTS isn’t supported.';
+
+  @override
+  String get ttsEngineServerDescription =>
+      'Always request audio from your OpenWebUI server.';
+
+  @override
+  String get ttsDeviceUnavailableWarning =>
+      'On-device text-to-speech isn’t available on this device.';
+
+  @override
+  String get ttsServerUnavailableWarning =>
+      'Connect to a server with text-to-speech enabled to use this option.';
+
  @override
  String get ttsSettings => 'Text to Speech';

--- a/lib/l10n/app_localizations_fr.dart
+++ b/lib/l10n/app_localizations_fr.dart
@@ -965,7 +965,7 @@ class AppLocalizationsFr extends AppLocalizations {

  @override
  String get sttEngineServerDescription =>
-      'Envoie toujours les enregistrements à votre serveur Conduit pour transcription.';
+      'Envoie toujours les enregistrements à votre serveur OpenWebUI pour transcription.';

  @override
  String get sttDeviceUnavailableWarning =>
@@ -978,12 +978,35 @@ class AppLocalizationsFr extends AppLocalizations {
  @override
  String get ttsEngineLabel => 'Moteur';

+  @override
+  String get ttsEngineAuto => 'Auto';
+
  @override
  String get ttsEngineDevice => 'Sur l\'appareil';

  @override
  String get ttsEngineServer => 'Serveur';

+  @override
+  String get ttsEngineAutoDescription =>
+      'Utilise la synthèse locale quand c’est possible, sinon bascule vers votre serveur.';
+
+  @override
+  String get ttsEngineDeviceDescription =>
+      'Garde la synthèse sur cet appareil. La lecture vocale ne fonctionne plus si l’appareil n’offre pas la synthèse vocale.';
+
+  @override
+  String get ttsEngineServerDescription =>
+      'Demande toujours l\'audio à votre serveur OpenWebUI.';
+
+  @override
+  String get ttsDeviceUnavailableWarning =>
+      'La synthèse vocale sur l’appareil n’est pas disponible sur cet appareil.';
+
+  @override
+  String get ttsServerUnavailableWarning =>
+      'Connectez-vous à un serveur avec la synthèse vocale activée pour utiliser cette option.';
+
  @override
  String get ttsSettings => 'Synthèse vocale';

--- a/lib/l10n/app_localizations_it.dart
+++ b/lib/l10n/app_localizations_it.dart
@@ -954,7 +954,7 @@ class AppLocalizationsIt extends AppLocalizations {

  @override
  String get sttEngineServerDescription =>
-      'Invia sempre le registrazioni al tuo server Conduit per la trascrizione.';
+      'Invia sempre le registrazioni al tuo server OpenWebUI per la trascrizione.';

  @override
  String get sttDeviceUnavailableWarning =>
@@ -967,12 +967,35 @@ class AppLocalizationsIt extends AppLocalizations {
  @override
  String get ttsEngineLabel => 'Motore';

+  @override
+  String get ttsEngineAuto => 'Automatico';
+
  @override
  String get ttsEngineDevice => 'Sul dispositivo';

  @override
  String get ttsEngineServer => 'Server';

+  @override
+  String get ttsEngineAutoDescription =>
+      'Usa la sintesi sul dispositivo quando disponibile e altrimenti passa al tuo server.';
+
+  @override
+  String get ttsEngineDeviceDescription =>
+      'Mantiene la sintesi su questo dispositivo. La riproduzione vocale non funziona se il dispositivo non supporta il TTS.';
+
+  @override
+  String get ttsEngineServerDescription =>
+      'Richiede sempre l\'audio dal tuo server OpenWebUI.';
+
+  @override
+  String get ttsDeviceUnavailableWarning =>
+      'La sintesi vocale sul dispositivo non è disponibile su questo dispositivo.';
+
+  @override
+  String get ttsServerUnavailableWarning =>
+      'Collegati a un server con la sintesi vocale abilitata per usare questa opzione.';
+
  @override
  String get ttsSettings => 'Sintesi vocale';

--- a/lib/l10n/app_nl.arb
+++ b/lib/l10n/app_nl.arb
@@ -314,10 +314,16 @@
  "sttEngineServer": "Server",
  "sttEngineAutoDescription": "Gebruikt spraakherkenning op het apparaat wanneer beschikbaar en valt anders terug op je server.",
  "sttEngineDeviceDescription": "Houdt audio op dit apparaat. Spraakinput werkt niet als het apparaat geen spraakherkenning ondersteunt.",
-  "sttEngineServerDescription": "Stuurt opnames altijd naar je Conduit-server voor transcriptie.",
+  "sttEngineServerDescription": "Stuurt opnames altijd naar je OpenWebUI-server voor transcriptie.",
  "sttDeviceUnavailableWarning": "Spraakherkenning op het apparaat is niet beschikbaar op dit apparaat.",
  "sttServerUnavailableWarning": "Verbind met een server met transcriptie ingeschakeld om deze optie te gebruiken.",
-  "ttsSettings": "Tekst naar spraak",
+"ttsSettings": "Tekst naar spraak",
+  "ttsEngineAuto": "Automatisch",
+  "ttsEngineAutoDescription": "Gebruikt spraaksynthese op het apparaat wanneer beschikbaar en valt anders terug op je server.",
+  "ttsEngineDeviceDescription": "Houdt de synthese op dit apparaat. Spraakweergave werkt niet als het apparaat geen TTS ondersteunt.",
+  "ttsEngineServerDescription": "Vraagt altijd audio op bij je OpenWebUI-server.",
+  "ttsDeviceUnavailableWarning": "Spraaksynthese op het apparaat is niet beschikbaar op dit apparaat.",
+  "ttsServerUnavailableWarning": "Verbind met een server met tekst-naar-spraak ingeschakeld om deze optie te gebruiken.",
  "ttsVoice": "Stem",
  "ttsSpeechRate": "Spraaksnelheid",
  "ttsPitch": "Toonhoogte",
--- a/lib/l10n/app_ru.arb
+++ b/lib/l10n/app_ru.arb
@@ -314,10 +314,16 @@
  "sttEngineServer": "Сервер",
  "sttEngineAutoDescription": "Использует распознавание на устройстве, когда это возможно, иначе переключается на ваш сервер.",
  "sttEngineDeviceDescription": "Оставляет звук на этом устройстве. Голосовой ввод не работает, если устройство не поддерживает распознавание речи.",
-  "sttEngineServerDescription": "Всегда отправляет записи на сервер Conduit для транскрибации.",
+  "sttEngineServerDescription": "Всегда отправляет записи на сервер OpenWebUI для транскрибации.",
  "sttDeviceUnavailableWarning": "Распознавание речи на устройстве недоступно на этом устройстве.",
  "sttServerUnavailableWarning": "Подключитесь к серверу с включённой транскрибацией, чтобы использовать эту опцию.",
-  "ttsSettings": "Преобразование текста в речь",
+"ttsSettings": "Преобразование текста в речь",
+  "ttsEngineAuto": "Авто",
+  "ttsEngineAutoDescription": "Использует синтез речи на устройстве, когда это возможно, иначе переключается на ваш сервер.",
+  "ttsEngineDeviceDescription": "Оставляет синтез на этом устройстве. Воспроизведение голоса не работает, если устройство не поддерживает синтез речи.",
+  "ttsEngineServerDescription": "Всегда запрашивает аудио у вашего сервера OpenWebUI.",
+  "ttsDeviceUnavailableWarning": "Синтез речи на устройстве недоступен на этом устройстве.",
+  "ttsServerUnavailableWarning": "Подключитесь к серверу с включённым синтезом речи, чтобы использовать эту опцию.",
  "ttsVoice": "Голос",
  "ttsSpeechRate": "Скорость речи",
  "ttsPitch": "Высота тона",
--- a/lib/l10n/app_zh.arb
+++ b/lib/l10n/app_zh.arb
@@ -314,10 +314,16 @@
  "sttEngineServer": "服务器",
  "sttEngineAutoDescription": "在可用时使用本机识别，否则切换到你的服务器。",
  "sttEngineDeviceDescription": "音频会保留在此设备上。如果设备不支持语音识别，语音输入将不可用。",
-  "sttEngineServerDescription": "始终将录音发送到你的 Conduit 服务器进行转写。",
+  "sttEngineServerDescription": "始终将录音发送到你的 OpenWebUI 服务器进行转写。",
  "sttDeviceUnavailableWarning": "此设备不支持本机语音识别。",
  "sttServerUnavailableWarning": "连接到启用转写功能的服务器后才能使用此选项。",
-  "ttsSettings": "文本转语音",
+"ttsSettings": "文本转语音",
+  "ttsEngineAuto": "自动",
+  "ttsEngineAutoDescription": "在可用时使用本机合成，否则切换到你的服务器。",
+  "ttsEngineDeviceDescription": "在此设备上完成合成。如果设备不支持文本转语音，语音播放将不可用。",
+  "ttsEngineServerDescription": "始终向你的 OpenWebUI 服务器请求音频。",
+  "ttsDeviceUnavailableWarning": "此设备不支持本机文本转语音。",
+  "ttsServerUnavailableWarning": "连接到启用文本转语音的服务器后才能使用此选项。",
  "ttsVoice": "语音",
  "ttsSpeechRate": "语速",
  "ttsPitch": "音调",