feat(soniox): add Soniox real-time streaming STT provider

main.js

@@ -1107,6 +1107,7 @@ if (gotSingleInstanceLock) {
       googleCalendarManager.stop();
     }
     if (ipcHandlers) {
+      ipcHandlers.cleanupAllStreaming();
       ipcHandlers._cleanupTextEditMonitor();
     }
     if (textEditMonitor) {

package.json

@@ -60,6 +60,7 @@
     "format:check": "eslint . && cd src && eslint . && prettier --check \"**/*.{js,jsx,ts,tsx,json,css,md}\"",
     "lint": "eslint . && cd src && eslint .",
     "typecheck": "cd src && tsc --noEmit",
+    "test": "node --test",
     "quality-check": "npm run format:check && npm run typecheck",
     "i18n:check": "node scripts/check-i18n.js",
     "preview": "cd src && vite preview",

preload.js

@@ -299,6 +299,10 @@ contextBridge.exposeInMainWorld("electronAPI", {
   saveMistralKey: (key) => ipcRenderer.invoke("save-mistral-key", key),
   proxyMistralTranscription: (data) => ipcRenderer.invoke("proxy-mistral-transcription", data),
 
+  // Soniox API
+  getSonioxKey: () => ipcRenderer.invoke("get-soniox-key"),
+  saveSonioxKey: (key) => ipcRenderer.invoke("save-soniox-key", key),
+
   // Custom endpoint API keys
   getCustomTranscriptionKey: () => ipcRenderer.invoke("get-custom-transcription-key"),
   saveCustomTranscriptionKey: (key) => ipcRenderer.invoke("save-custom-transcription-key", key),
@@ -493,6 +497,30 @@ contextBridge.exposeInMainWorld("electronAPI", {
     (callback) => (_event, data) => callback(data)
   ),
 
+  // Soniox streaming
+  sonioxStreamingWarmup: (options) => ipcRenderer.invoke("soniox-streaming-warmup", options),
+  sonioxStreamingStart: (options) => ipcRenderer.invoke("soniox-streaming-start", options),
+  sonioxStreamingSend: (audioBuffer) => ipcRenderer.send("soniox-streaming-send", audioBuffer),
+  sonioxStreamingFinalize: () => ipcRenderer.send("soniox-streaming-finalize"),
+  sonioxStreamingStop: () => ipcRenderer.invoke("soniox-streaming-stop"),
+  sonioxStreamingStatus: () => ipcRenderer.invoke("soniox-streaming-status"),
+  onSonioxPartialTranscript: registerListener(
+    "soniox-streaming-partial",
+    (callback) => (_event, text) => callback(text)
+  ),
+  onSonioxFinalTranscript: registerListener(
+    "soniox-streaming-final",
+    (callback) => (_event, text) => callback(text)
+  ),
+  onSonioxError: registerListener(
+    "soniox-streaming-error",
+    (callback) => (_event, error) => callback(error)
+  ),
+  onSonioxSessionEnd: registerListener(
+    "soniox-streaming-session-end",
+    (callback) => (_event, data) => callback(data)
+  ),
+
   // Usage limit events (for showing UpgradePrompt in ControlPanel)
   notifyLimitReached: (data) => ipcRenderer.send("limit-reached", data),
   onLimitReached: registerListener("limit-reached", (callback) => (_event, data) => callback(data)),

src/components/OnboardingFlow.tsx

@@ -83,6 +83,8 @@ export default function OnboardingFlow({ onComplete }: OnboardingFlowProps) {
     openaiApiKey,
     groqApiKey,
     mistralApiKey,
+    sonioxApiKey,
+    setSonioxApiKey,
     customTranscriptionApiKey,
     setCustomTranscriptionApiKey,
     dictationKey,
@@ -502,6 +504,8 @@ export default function OnboardingFlow({ onComplete }: OnboardingFlowProps) {
               setGroqApiKey={setGroqApiKey}
               mistralApiKey={mistralApiKey}
               setMistralApiKey={setMistralApiKey}
+              sonioxApiKey={sonioxApiKey}
+              setSonioxApiKey={setSonioxApiKey}
               customTranscriptionApiKey={customTranscriptionApiKey}
               setCustomTranscriptionApiKey={setCustomTranscriptionApiKey}
               cloudTranscriptionBaseUrl={cloudTranscriptionBaseUrl}
@@ -725,6 +729,8 @@ export default function OnboardingFlow({ onComplete }: OnboardingFlowProps) {
             return groqApiKey.trim().length > 0;
           } else if (cloudTranscriptionProvider === "mistral") {
             return mistralApiKey.trim().length > 0;
+          } else if (cloudTranscriptionProvider === "soniox") {
+            return sonioxApiKey.trim().length > 0;
           } else if (cloudTranscriptionProvider === "custom") {
             // Custom can work without API key for local endpoints
             return true;

src/components/SettingsPage.tsx

@@ -172,6 +172,10 @@ interface TranscriptionSectionProps {
   setGroqApiKey: (key: string) => void;
   mistralApiKey: string;
   setMistralApiKey: (key: string) => void;
+  sonioxApiKey: string;
+  setSonioxApiKey: (key: string) => void;
+  sonioxSecondaryLanguage: string;
+  setSonioxSecondaryLanguage: (lang: string) => void;
   customTranscriptionApiKey: string;
   setCustomTranscriptionApiKey: (key: string) => void;
   cloudTranscriptionBaseUrl?: string;
@@ -207,6 +211,10 @@ function TranscriptionSection({
   setGroqApiKey,
   mistralApiKey,
   setMistralApiKey,
+  sonioxApiKey,
+  setSonioxApiKey,
+  sonioxSecondaryLanguage,
+  setSonioxSecondaryLanguage,
   customTranscriptionApiKey,
   setCustomTranscriptionApiKey,
   cloudTranscriptionBaseUrl,
@@ -383,6 +391,10 @@ function TranscriptionSection({
           setGroqApiKey={setGroqApiKey}
           mistralApiKey={mistralApiKey}
           setMistralApiKey={setMistralApiKey}
+          sonioxApiKey={sonioxApiKey}
+          setSonioxApiKey={setSonioxApiKey}
+          sonioxSecondaryLanguage={sonioxSecondaryLanguage}
+          setSonioxSecondaryLanguage={setSonioxSecondaryLanguage}
           customTranscriptionApiKey={customTranscriptionApiKey}
           setCustomTranscriptionApiKey={setCustomTranscriptionApiKey}
           cloudTranscriptionBaseUrl={cloudTranscriptionBaseUrl}
@@ -676,6 +688,10 @@ export default function SettingsPage({ activeSection = "general" }: SettingsPage
     setGeminiApiKey,
     setGroqApiKey,
     setMistralApiKey,
+    sonioxApiKey,
+    setSonioxApiKey,
+    sonioxSecondaryLanguage,
+    setSonioxSecondaryLanguage,
     customTranscriptionApiKey,
     setCustomTranscriptionApiKey,
     customReasoningApiKey,
@@ -2670,6 +2686,10 @@ EOF`,
             setGroqApiKey={setGroqApiKey}
             mistralApiKey={mistralApiKey}
             setMistralApiKey={setMistralApiKey}
+            sonioxApiKey={sonioxApiKey}
+            setSonioxApiKey={setSonioxApiKey}
+            sonioxSecondaryLanguage={sonioxSecondaryLanguage}
+            setSonioxSecondaryLanguage={setSonioxSecondaryLanguage}
             customTranscriptionApiKey={customTranscriptionApiKey}
             setCustomTranscriptionApiKey={setCustomTranscriptionApiKey}
             cloudTranscriptionBaseUrl={cloudTranscriptionBaseUrl}

src/components/TranscriptionModelPicker.tsx

@@ -8,6 +8,8 @@ import { ProviderTabs } from "./ui/ProviderTabs";
 import ModelCardList from "./ui/ModelCardList";
 import { DownloadProgressBar } from "./ui/DownloadProgressBar";
 import ApiKeyInput from "./ui/ApiKeyInput";
+import LanguageSelector, { type LanguageOption } from "./ui/LanguageSelector";
+import languageRegistry from "../config/languageRegistry.json";
 import { ConfirmDialog } from "./ui/dialog";
 import { useDialogs } from "../hooks/useDialogs";
 import { useModelDownload, type DownloadProgress } from "../hooks/useModelDownload";
@@ -26,6 +28,7 @@ import { getProviderIcon, isMonochromeProvider } from "../utils/providerIcons";
 import { API_ENDPOINTS, normalizeBaseUrl } from "../config/constants";
 import { createExternalLinkHandler } from "../utils/externalLinks";
 import { getCachedPlatform } from "../utils/platform";
+import { useSettingsStore } from "../stores/settingsStore";
 import type { CudaWhisperStatus } from "../types/electron";
 import logger from "../utils/logger";
 
@@ -199,16 +202,25 @@ interface TranscriptionModelPickerProps {
   setMistralApiKey: (key: string) => void;
   customTranscriptionApiKey?: string;
   setCustomTranscriptionApiKey?: (key: string) => void;
+  sonioxApiKey?: string;
+  setSonioxApiKey?: (key: string) => void;
+  sonioxSecondaryLanguage?: string;
+  setSonioxSecondaryLanguage?: (lang: string) => void;
   cloudTranscriptionBaseUrl?: string;
   setCloudTranscriptionBaseUrl?: (url: string) => void;
   className?: string;
   variant?: "onboarding" | "settings";
 }
 
+const SECONDARY_LANGUAGE_OPTIONS: LanguageOption[] = languageRegistry.languages
+  .filter((l) => l.code !== "auto")
+  .map(({ code, label, flag }) => ({ value: code, label, flag }));
+
 const CLOUD_PROVIDER_TABS = [
   { id: "openai", name: "OpenAI" },
   { id: "groq", name: "Groq", recommended: true },
   { id: "mistral", name: "Mistral" },
+  { id: "soniox", name: "Soniox" },
   { id: "custom", name: "Custom" },
 ];
 
@@ -274,12 +286,18 @@ export default function TranscriptionModelPicker({
   setMistralApiKey,
   customTranscriptionApiKey = "",
   setCustomTranscriptionApiKey,
+  sonioxApiKey = "",
+  setSonioxApiKey,
+  sonioxSecondaryLanguage = "",
+  setSonioxSecondaryLanguage,
   cloudTranscriptionBaseUrl = "",
   setCloudTranscriptionBaseUrl,
   className = "",
   variant = "settings",
 }: TranscriptionModelPickerProps) {
   const { t } = useTranslation();
+  const preferredLanguage = useSettingsStore((s) => s.preferredLanguage);
+  const isAutoLanguage = !preferredLanguage || preferredLanguage === "auto";
   const [localModels, setLocalModels] = useState<LocalModel[]>([]);
   const [parakeetModels, setParakeetModels] = useState<LocalModel[]>([]);
   const [internalLocalProvider, setInternalLocalProvider] = useState(selectedLocalProvider);
@@ -393,8 +411,13 @@ export default function TranscriptionModelPicker({
           }
         }
       }
-    } else if (selectedCloudProvider !== "custom" && !selectedCloudModel) {
-      const provider = cloudProviders.find((p) => p.id === selectedCloudProvider);
+    } else if (
+      selectedCloudProvider !== "custom" &&
+      !selectedCloudModel
+    ) {
+      const provider = cloudProviders.find(
+        (p) => p.id === selectedCloudProvider
+      );
       if (provider?.models?.length) {
         onCloudModelSelect(provider.models[0].id);
       }
@@ -527,7 +550,9 @@ export default function TranscriptionModelPicker({
   const handleCloudProviderChange = useCallback(
     (providerId: string) => {
       onCloudProviderSelect(providerId);
-      const provider = cloudProviders.find((p) => p.id === providerId);
+      const provider = cloudProviders.find(
+        (p) => p.id === providerId
+      );
 
       if (providerId === "custom") {
         onCloudModelSelect("whisper-1");
@@ -541,7 +566,12 @@ export default function TranscriptionModelPicker({
         }
       }
     },
-    [cloudProviders, onCloudProviderSelect, onCloudModelSelect, setCloudTranscriptionBaseUrl]
+    [
+      cloudProviders,
+      onCloudProviderSelect,
+      onCloudModelSelect,
+      setCloudTranscriptionBaseUrl,
+    ]
   );
 
   const handleLocalProviderChange = useCallback(
@@ -837,7 +867,11 @@ export default function TranscriptionModelPicker({
                   </label>
                   <Input
                     value={cloudTranscriptionBaseUrl}
-                    onChange={(e) => setCloudTranscriptionBaseUrl?.(e.target.value)}
+                    onChange={(e) =>
+                      setCloudTranscriptionBaseUrl?.(
+                        e.target.value
+                      )
+                    }
                     onBlur={handleBaseUrlBlur}
                     placeholder="https://your-api.example.com/v1"
                     className="h-8 text-sm"
@@ -846,7 +880,10 @@ export default function TranscriptionModelPicker({
 
                 <ApiKeyInput
                   apiKey={customTranscriptionApiKey}
-                  setApiKey={setCustomTranscriptionApiKey || (() => {})}
+                  setApiKey={
+                    setCustomTranscriptionApiKey ||
+                    (() => {})
+                  }
                   label={t("transcription.apiKeyOptional")}
                   helpText=""
                 />
@@ -857,7 +894,9 @@ export default function TranscriptionModelPicker({
                   </label>
                   <Input
                     value={selectedCloudModel}
-                    onChange={(e) => onCloudModelSelect(e.target.value)}
+                    onChange={(e) =>
+                      onCloudModelSelect(e.target.value)
+                    }
                     placeholder="whisper-1"
                     className="h-8 text-sm"
                   />
@@ -877,7 +916,9 @@ export default function TranscriptionModelPicker({
                           groq: "https://console.groq.com/keys",
                           mistral: "https://console.mistral.ai/api-keys",
                           openai: "https://platform.openai.com/api-keys",
-                        }[selectedCloudProvider] || "https://platform.openai.com/api-keys"
+                          soniox: "https://console.soniox.com/",
+                        }[selectedCloudProvider] ||
+                          "https://platform.openai.com/api-keys"
                       )}
                       className="text-xs text-primary/70 hover:text-primary transition-colors cursor-pointer"
                     >
@@ -886,22 +927,49 @@ export default function TranscriptionModelPicker({
                   </div>
                   <ApiKeyInput
                     apiKey={
-                      { groq: groqApiKey, mistral: mistralApiKey, openai: openaiApiKey }[
-                        selectedCloudProvider
-                      ] || openaiApiKey
+                      {
+                        groq: groqApiKey,
+                        mistral: mistralApiKey,
+                        openai: openaiApiKey,
+                        soniox: sonioxApiKey,
+                      }[selectedCloudProvider] ||
+                      openaiApiKey
                     }
                     setApiKey={
-                      { groq: setGroqApiKey, mistral: setMistralApiKey, openai: setOpenaiApiKey }[
-                        selectedCloudProvider
-                      ] || setOpenaiApiKey
+                      {
+                        groq: setGroqApiKey,
+                        mistral: setMistralApiKey,
+                        openai: setOpenaiApiKey,
+                        soniox: setSonioxApiKey,
+                      }[selectedCloudProvider] ||
+                      setOpenaiApiKey
                     }
                     label=""
                     helpText=""
                   />
                 </div>
 
+                {selectedCloudProvider === "soniox" && setSonioxSecondaryLanguage && (
+                  <div className={`flex items-center justify-between gap-3 ${isAutoLanguage ? "opacity-50 pointer-events-none" : ""}`}>
+                    <label className="text-xs font-medium text-foreground whitespace-nowrap">
+                      {t("common.secondaryLanguage")}
+                    </label>
+                    <LanguageSelector
+                      value={isAutoLanguage ? "none" : (sonioxSecondaryLanguage || "none")}
+                      onChange={(value) => setSonioxSecondaryLanguage(value === "none" ? "" : value)}
+                      options={[
+                        { value: "none", label: t("common.none"), flag: "" },
+                        ...SECONDARY_LANGUAGE_OPTIONS,
+                      ]}
+                      className="min-w-32"
+                    />
+                  </div>
+                )}
+
                 <div className="space-y-1.5">
-                  <label className="text-xs font-medium text-foreground">{t("common.model")}</label>
+                  <label className="text-xs font-medium text-foreground">
+                    {t("common.model")}
+                  </label>
                   <ModelCardList
                     models={cloudModelOptions}
                     selectedModel={selectedCloudModel}