diff --git a/src/components/SettingsPage.tsx b/src/components/SettingsPage.tsx
index 9178d5bc..9e1fe89e 100644
--- a/src/components/SettingsPage.tsx
+++ b/src/components/SettingsPage.tsx
@@ -26,6 +26,7 @@ import {
   RotateCw,
   BookOpen,
   Copy,
+  Zap,
 } from "lucide-react";
 import { useAuth } from "../hooks/useAuth";
 import { NEON_AUTH_URL, signOut } from "../lib/neonAuth";
@@ -176,6 +177,8 @@ interface TranscriptionSectionProps {
   setCustomTranscriptionApiKey: (key: string) => void;
   cloudTranscriptionBaseUrl?: string;
   setCloudTranscriptionBaseUrl: (url: string) => void;
+  dictationMode: "normal" | "fast";
+  setDictationMode: (mode: string) => void;
   toast: (opts: {
     title: string;
     description: string;
@@ -211,6 +214,8 @@ function TranscriptionSection({
   setCustomTranscriptionApiKey,
   cloudTranscriptionBaseUrl,
   setCloudTranscriptionBaseUrl,
+  dictationMode,
+  setDictationMode,
   toast,
 }: TranscriptionSectionProps) {
   const { t } = useTranslation();
@@ -350,6 +355,57 @@ function TranscriptionSection({
         </SettingsPanel>
       )}
 
+      {/* Dictation mode selector — visible for all cloud modes, signed in or not */}
+      {!useLocalWhisper && (
+        <SettingsPanel>
+          <SettingsPanelRow>
+            <p className="text-xs font-medium text-muted-foreground/80 mb-2">
+              {t("settingsPage.transcription.dictationMode.title")}
+            </p>
+            <div className="flex gap-2">
+              <button
+                onClick={() => {
+                  setDictationMode("normal");
+                  logger.info("Dictation mode changed", { mode: "batch" }, "audio");
+                }}
+                className={`flex-1 flex items-center gap-2 px-3 py-2 rounded-md border text-xs font-medium transition-colors cursor-pointer ${
+                  dictationMode === "normal"
+                    ? "border-primary bg-primary/10 dark:bg-primary/15 text-primary"
+                    : "border-border bg-background text-muted-foreground hover:bg-muted/60"
+                }`}
+              >
+                <Mic className="w-3.5 h-3.5 shrink-0" />
+                <div className="text-left">
+                  <div>{t("settingsPage.transcription.dictationMode.normal")}</div>
+                  <div className="font-normal text-muted-foreground/70 mt-0.5">
+                    {t("settingsPage.transcription.dictationMode.normalDescription")}
+                  </div>
+                </div>
+              </button>
+              <button
+                onClick={() => {
+                  setDictationMode("fast");
+                  logger.info("Dictation mode changed", { mode: "streaming" }, "streaming");
+                }}
+                className={`flex-1 flex items-center gap-2 px-3 py-2 rounded-md border text-xs font-medium transition-colors cursor-pointer ${
+                  dictationMode === "fast"
+                    ? "border-primary bg-primary/10 dark:bg-primary/15 text-primary"
+                    : "border-border bg-background text-muted-foreground hover:bg-muted/60"
+                }`}
+              >
+                <Zap className="w-3.5 h-3.5 shrink-0" />
+                <div className="text-left">
+                  <div>{t("settingsPage.transcription.dictationMode.fast")}</div>
+                  <div className="font-normal text-muted-foreground/70 mt-0.5">
+                    {t("settingsPage.transcription.dictationMode.fastDescription")}
+                  </div>
+                </div>
+              </button>
+            </div>
+          </SettingsPanelRow>
+        </SettingsPanel>
+      )}
+
       {/* Custom Setup model picker — shown when Custom Setup is active or not signed in */}
       {(isCustomMode || !isSignedIn) && (
         <TranscriptionModelPicker
@@ -709,6 +765,8 @@ export default function SettingsPage({ activeSection = "general" }: SettingsPage
     setAudioRetentionDays,
     customDictionary,
     setCustomDictionary,
+    dictationMode,
+    setDictationMode,
   } = useSettings();
 
   const { t, i18n } = useTranslation();
@@ -2674,6 +2732,8 @@ EOF`,
             setCustomTranscriptionApiKey={setCustomTranscriptionApiKey}
             cloudTranscriptionBaseUrl={cloudTranscriptionBaseUrl}
             setCloudTranscriptionBaseUrl={setCloudTranscriptionBaseUrl}
+            dictationMode={dictationMode}
+            setDictationMode={setDictationMode}
             toast={toast}
           />
         );
diff --git a/src/helpers/audioManager.js b/src/helpers/audioManager.js
index 4d49b91f..2f24db78 100644
--- a/src/helpers/audioManager.js
+++ b/src/helpers/audioManager.js
@@ -324,6 +324,13 @@ registerProcessor("pcm-streaming-processor", PCMStreamingProcessor);
         );
       }
 
+      const useStreaming = this.shouldUseStreaming();
+      logger.info(
+        "Transcription mode",
+        { mode: useStreaming ? "streaming" : "batch" },
+        useStreaming ? "streaming" : "audio"
+      );
+
       // Mix in system audio if enabled
       let recordingStream = micStream;
       if (this.systemAudioEnabled) {
@@ -2003,6 +2010,7 @@ registerProcessor("pcm-streaming-processor", PCMStreamingProcessor);
 
   shouldUseStreaming(isSignedInOverride) {
     const s = getSettings();
+    if (s.dictationMode === "normal") return false;
     if (s.useLocalWhisper) return false;
 
     if (REALTIME_MODELS.has(s.cloudTranscriptionModel)) {
diff --git a/src/hooks/useSettings.ts b/src/hooks/useSettings.ts
index 56d0c62f..a91070b0 100644
--- a/src/hooks/useSettings.ts
+++ b/src/hooks/useSettings.ts
@@ -20,6 +20,7 @@ export interface TranscriptionSettings {
   cloudTranscriptionMode: string;
   customDictionary: string[];
   assemblyAiStreaming: boolean;
+  dictationMode: "normal" | "fast";
 }
 
 export interface ReasoningSettings {
@@ -179,6 +180,8 @@ function useSettingsInternal() {
     customDictionary: store.customDictionary,
     assemblyAiStreaming: store.assemblyAiStreaming,
     setAssemblyAiStreaming: store.setAssemblyAiStreaming,
+    dictationMode: store.dictationMode,
+    setDictationMode: store.setDictationMode,
     useReasoningModel: store.useReasoningModel,
     reasoningModel: store.reasoningModel,
     reasoningProvider: store.reasoningProvider,
diff --git a/src/locales/de/translation.json b/src/locales/de/translation.json
index 3bb3a6dd..6667ff9b 100644
--- a/src/locales/de/translation.json
+++ b/src/locales/de/translation.json
@@ -1319,6 +1319,13 @@
           "description": "Konfigurieren Sie Ihren eigenen Anbieter und API-Schlüssel.",
           "title": "Zu benutzerdefinierter Einrichtung gewechselt"
         }
+      },
+      "dictationMode": {
+        "title": "Diktiermodus",
+        "normal": "Normal",
+        "normalDescription": "Beste Genauigkeit, braucht aber einen Moment zur Verarbeitung nach der Aufnahme.",
+        "fast": "Schnell",
+        "fastDescription": "WebSocket-Streaming über OpenAI Realtime API. Live-Transkription während Sie sprechen."
       }
     }
   },
diff --git a/src/locales/en/translation.json b/src/locales/en/translation.json
index 9cfa4475..f55e08f0 100644
--- a/src/locales/en/translation.json
+++ b/src/locales/en/translation.json
@@ -1389,6 +1389,13 @@
           "description": "Configure your own provider and API key.",
           "title": "Switched to Custom Setup"
         }
+      },
+      "dictationMode": {
+        "title": "Dictation Mode",
+        "normal": "Normal",
+        "normalDescription": "Best accuracy, but takes a moment to process after recording.",
+        "fast": "Fast",
+        "fastDescription": "WebSocket streaming via OpenAI Realtime API. Live transcription as you speak."
       }
     }
   },
diff --git a/src/locales/es/translation.json b/src/locales/es/translation.json
index 65c8f091..41e22639 100644
--- a/src/locales/es/translation.json
+++ b/src/locales/es/translation.json
@@ -1319,6 +1319,13 @@
           "description": "Configura tu propio proveedor y clave API.",
           "title": "Cambiado a configuración personalizada"
         }
+      },
+      "dictationMode": {
+        "title": "Modo de dictado",
+        "normal": "Normal",
+        "normalDescription": "Máxima precisión, pero tarda un momento en procesar tras la grabación.",
+        "fast": "Rápido",
+        "fastDescription": "Streaming WebSocket a través de OpenAI Realtime API. Transcripción en tiempo real."
       }
     }
   },
diff --git a/src/locales/fr/translation.json b/src/locales/fr/translation.json
index 50b0ef89..698ca097 100644
--- a/src/locales/fr/translation.json
+++ b/src/locales/fr/translation.json
@@ -1319,6 +1319,13 @@
           "description": "Configurez votre propre fournisseur et clé API.",
           "title": "Basculé vers la configuration personnalisée"
         }
+      },
+      "dictationMode": {
+        "title": "Mode de dictée",
+        "normal": "Normal",
+        "normalDescription": "Meilleure précision, mais prend un moment pour traiter après l'enregistrement.",
+        "fast": "Rapide",
+        "fastDescription": "Streaming WebSocket via OpenAI Realtime API. Transcription en temps réel."
       }
     }
   },
diff --git a/src/locales/it/translation.json b/src/locales/it/translation.json
index 5577e040..b6fbc462 100644
--- a/src/locales/it/translation.json
+++ b/src/locales/it/translation.json
@@ -1319,6 +1319,13 @@
           "description": "Configura il tuo provider e la tua chiave API.",
           "title": "Passato a configurazione personalizzata"
         }
+      },
+      "dictationMode": {
+        "title": "Modalità dettatura",
+        "normal": "Normale",
+        "normalDescription": "Massima accuratezza, ma richiede un momento per elaborare dopo la registrazione.",
+        "fast": "Veloce",
+        "fastDescription": "Streaming WebSocket tramite OpenAI Realtime API. Trascrizione in tempo reale."
       }
     }
   },
diff --git a/src/locales/ja/translation.json b/src/locales/ja/translation.json
index 6584d4fe..1b964f3d 100644
--- a/src/locales/ja/translation.json
+++ b/src/locales/ja/translation.json
@@ -1319,6 +1319,13 @@
           "description": "独自のプロバイダーと API キーを設定してください。",
           "title": "カスタム設定に切り替えました"
         }
+      },
+      "dictationMode": {
+        "title": "ディクテーションモード",
+        "normal": "通常",
+        "normalDescription": "最高の精度ですが、録音後に処理に少し時間がかかります。",
+        "fast": "高速",
+        "fastDescription": "OpenAI Realtime API経由のWebSocketストリーミング。リアルタイム文字起こし。"
       }
     }
   },
diff --git a/src/locales/pt/translation.json b/src/locales/pt/translation.json
index d49d8e62..72fbdbff 100644
--- a/src/locales/pt/translation.json
+++ b/src/locales/pt/translation.json
@@ -1291,6 +1291,13 @@
           "description": "Configure seu próprio provedor e chave API.",
           "title": "Mudou para configuração personalizada"
         }
+      },
+      "dictationMode": {
+        "title": "Modo de ditado",
+        "normal": "Normal",
+        "normalDescription": "Melhor precisão, mas leva um momento para processar após a gravação.",
+        "fast": "Rápido",
+        "fastDescription": "Streaming WebSocket via OpenAI Realtime API. Transcrição em tempo real."
       }
     }
   },
diff --git a/src/locales/ru/translation.json b/src/locales/ru/translation.json
index 0f98f327..430fffb3 100644
--- a/src/locales/ru/translation.json
+++ b/src/locales/ru/translation.json
@@ -1319,6 +1319,13 @@
           "description": "Настройте собственного провайдера и API-ключ.",
           "title": "Переключено на пользовательскую настройку"
         }
+      },
+      "dictationMode": {
+        "title": "Режим диктовки",
+        "normal": "Обычный",
+        "normalDescription": "Лучшая точность, но требуется момент на обработку после записи.",
+        "fast": "Быстрый",
+        "fastDescription": "WebSocket-стриминг через OpenAI Realtime API. Транскрипция в реальном времени."
       }
     }
   },
diff --git a/src/locales/zh-CN/translation.json b/src/locales/zh-CN/translation.json
index b12e1c62..2757b379 100644
--- a/src/locales/zh-CN/translation.json
+++ b/src/locales/zh-CN/translation.json
@@ -1319,6 +1319,13 @@
           "description": "请配置你的服务商和 API Key。",
           "title": "已切换到自定义设置"
         }
+      },
+      "dictationMode": {
+        "title": "听写模式",
+        "normal": "普通",
+        "normalDescription": "最佳准确度，但录音后需要片刻处理。",
+        "fast": "快速",
+        "fastDescription": "通过 OpenAI Realtime API 进行 WebSocket 流式传输。实时转录。"
       }
     }
   },
diff --git a/src/locales/zh-TW/translation.json b/src/locales/zh-TW/translation.json
index 80c45443..d5ef1540 100644
--- a/src/locales/zh-TW/translation.json
+++ b/src/locales/zh-TW/translation.json
@@ -1319,6 +1319,13 @@
           "description": "設定你自己的供應商和 API Key。",
           "title": "已切換至自訂設定"
         }
+      },
+      "dictationMode": {
+        "title": "聽寫模式",
+        "normal": "一般",
+        "normalDescription": "最佳準確度，但錄音後需要片刻處理。",
+        "fast": "快速",
+        "fastDescription": "透過 OpenAI Realtime API 進行 WebSocket 串流。即時轉錄。"
       }
     }
   },
diff --git a/src/stores/settingsStore.ts b/src/stores/settingsStore.ts
index bfb84313..87a34418 100644
--- a/src/stores/settingsStore.ts
+++ b/src/stores/settingsStore.ts
@@ -121,6 +121,7 @@ export interface SettingsState
   setCloudReasoningBaseUrl: (value: string) => void;
   setCustomDictionary: (words: string[]) => void;
   setAssemblyAiStreaming: (value: boolean) => void;
+  setDictationMode: (value: string) => void;
   setUseReasoningModel: (value: boolean) => void;
   setReasoningModel: (value: string) => void;
   setReasoningProvider: (value: string) => void;
@@ -242,6 +243,9 @@ export const useSettingsStore = create<SettingsState>()((set, get) => ({
   cloudReasoningBaseUrl: readString("cloudReasoningBaseUrl", API_ENDPOINTS.OPENAI_BASE),
   customDictionary: readStringArray("customDictionary", []),
   assemblyAiStreaming: readBoolean("assemblyAiStreaming", true),
+  dictationMode: (readString("dictationMode", "normal") === "fast" ? "fast" : "normal") as
+    | "normal"
+    | "fast",
 
   useReasoningModel: readBoolean("useReasoningModel", true),
   reasoningModel: readString("reasoningModel", ""),
@@ -330,6 +334,7 @@ export const useSettingsStore = create<SettingsState>()((set, get) => ({
   setCloudReasoningMode: createStringSetter("cloudReasoningMode"),
   setCloudReasoningBaseUrl: createStringSetter("cloudReasoningBaseUrl"),
   setAssemblyAiStreaming: createBooleanSetter("assemblyAiStreaming"),
+  setDictationMode: createStringSetter("dictationMode"),
   setUseReasoningModel: createBooleanSetter("useReasoningModel"),
   setReasoningModel: createStringSetter("reasoningModel"),
   setReasoningProvider: createStringSetter("reasoningProvider"),
@@ -524,6 +529,7 @@ export const useSettingsStore = create<SettingsState>()((set, get) => ({
     if (settings.customDictionary !== undefined) s.setCustomDictionary(settings.customDictionary);
     if (settings.assemblyAiStreaming !== undefined)
       s.setAssemblyAiStreaming(settings.assemblyAiStreaming);
+    if (settings.dictationMode !== undefined) s.setDictationMode(settings.dictationMode);
   },
 
   updateReasoningSettings: (settings: Partial<ReasoningSettings>) => {