feat: surface all xAI TTS params in desktop GUI config

- Add speed, auto_speech_tags, text_normalization,
  optimize_streaming_latency, sample_rate, bit_rate to
  DEFAULT_CONFIG tts.xai block (backend schema source)
- Add field labels, descriptions, and section keys in
  frontend constants.ts for all 7 xAI TTS fields
- Update i18n translations (ja, zh, zh-hant)
- Fix stale tts.provider options in web_server.py schema
  overrides (was missing xai, minimax, mistral, gemini,
  kittentts, piper)

apps/desktop/src/app/settings/constants.ts

@@ -384,7 +384,13 @@ export const FIELD_LABELS: Record<string, string> = defineFieldCopy({
     },
     xai: {
       voiceId: 'xAI (Grok) Voice',
-      language: 'xAI Language'
+      language: 'xAI Language',
+      speed: 'Playback Speed',
+      autoSpeechTags: 'Auto Speech Tags',
+      textNormalization: 'Text Normalization',
+      optimizeStreamingLatency: 'Streaming Latency Optimization',
+      sampleRate: 'Sample Rate',
+      bitRate: 'Bit Rate'
     },
     minimax: {
       model: 'MiniMax TTS Model',
@@ -491,7 +497,13 @@ export const FIELD_DESCRIPTIONS: Record<string, string> = defineFieldCopy({
   tts: {
     xai: {
       voiceId: 'xAI voice ID (e.g. eve) or a custom voice ID.',
-      language: 'Spoken language code, e.g. en.'
+      language: 'Spoken language code (e.g. en, pt-BR) or "auto" for auto-detection.',
+      speed: 'Playback speed. 0.7 = slower, 1.0 = normal, 1.5 = faster.',
+      autoSpeechTags: 'Let an LLM insert expressive audio tags ([laughing], [sighs]) into the script before synthesis.',
+      textNormalization: 'Convert numbers, abbreviations, and symbols to spoken form before synthesis.',
+      optimizeStreamingLatency: 'Latency vs. quality trade-off. 0 = best quality, 2 = lowest latency.',
+      sampleRate: 'Audio sample rate in Hz. Higher = better quality, larger files.',
+      bitRate: 'MP3 bitrate in bps. Only applies when codec is mp3.'
     },
     neutts: {
       device: 'Local inference device for NeuTTS.'
@@ -592,6 +604,12 @@ export const SECTIONS: DesktopConfigSection[] = [
       'tts.elevenlabs.model_id',
       'tts.xai.voice_id',
       'tts.xai.language',
+      'tts.xai.speed',
+      'tts.xai.auto_speech_tags',
+      'tts.xai.text_normalization',
+      'tts.xai.optimize_streaming_latency',
+      'tts.xai.sample_rate',
+      'tts.xai.bit_rate',
       'tts.minimax.model',
       'tts.minimax.voice_id',
       'tts.mistral.model',

apps/desktop/src/i18n/ja.ts

@@ -465,7 +465,13 @@ export const ja = defineLocale({
         },
         xai: {
           voiceId: 'xAI (Grok) 音声',
-          language: 'xAI 言語'
+          language: 'xAI 言語',
+          speed: '再生速度',
+          autoSpeechTags: '自動音声タグ',
+          textNormalization: 'テキスト正規化',
+          optimizeStreamingLatency: 'ストリーミング遅延最適化',
+          sampleRate: 'サンプルレート',
+          bitRate: 'ビットレート'
         },
         minimax: {
           model: 'MiniMax TTS モデル',

apps/desktop/src/i18n/zh-hant.ts

@@ -454,7 +454,13 @@ export const zhHant = defineLocale({
         },
         xai: {
           voiceId: 'xAI (Grok) 語音',
-          language: 'xAI 語言'
+          language: 'xAI 語言',
+          speed: '播放速度',
+          autoSpeechTags: '自動語音標籤',
+          textNormalization: '文字正規化',
+          optimizeStreamingLatency: '串流延遲最佳化',
+          sampleRate: '取樣率',
+          bitRate: '位元率'
         },
         minimax: {
           model: 'MiniMax TTS 模型',

apps/desktop/src/i18n/zh.ts

@@ -565,7 +565,13 @@ export const zh: Translations = {
         },
         xai: {
           voiceId: 'xAI (Grok) 语音',
-          language: 'xAI 语言'
+          language: 'xAI 语言',
+          speed: '播放速度',
+          autoSpeechTags: '自动语音标签',
+          textNormalization: '文本规范化',
+          optimizeStreamingLatency: '流式延迟优化',
+          sampleRate: '采样率',
+          bitRate: '比特率'
         },
         minimax: {
           model: 'MiniMax TTS 模型',

hermes_cli/config.py

@@ -2184,9 +2184,13 @@ DEFAULT_CONFIG = {
         },
         "xai": {
             "voice_id": "eve",  # or custom voice ID — see https://docs.x.ai/developers/model-capabilities/audio/custom-voices
-            "language": "en",
-            "sample_rate": 24000,
-            "bit_rate": 128000,
+            "language": "en",  # BCP-47 code ("en", "pt-BR") or "auto"
+            "speed": 1.0,  # 0.7–1.5, playback speed
+            "auto_speech_tags": False,  # insert expressive audio tags via LLM rewrite
+            "text_normalization": False,  # normalize numbers/abbreviations/symbols to spoken form
+            "optimize_streaming_latency": 0,  # 0–2, trades quality for lower latency
+            "sample_rate": 24000,  # 22050 / 24000 / 44100 / 48000
+            "bit_rate": 128000,  # MP3 bitrate; only applies when codec=mp3
         },
         "mistral": {
             "model": "voxtral-mini-tts-2603",

hermes_cli/web_server.py

@@ -668,7 +668,7 @@ _SCHEMA_OVERRIDES: Dict[str, Dict[str, Any]] = {
     "tts.provider": {
         "type": "select",
         "description": "Text-to-speech provider",
-        "options": ["edge", "elevenlabs", "openai", "neutts"],
+        "options": ["edge", "elevenlabs", "openai", "xai", "minimax", "mistral", "gemini", "neutts", "kittentts", "piper"],
     },
     "stt.provider": {
         "type": "select",