Merge pull request #128 from AssemblyAI/D0068A307C832CBB8356B9DDF54BB16A

Sync from internal repo (YYYY/MM/DD)

Author: zkleb-aai

CHANGELOG.md

@@ -22,7 +22,7 @@
 
 - Add `language_confidence_threshold` to `Transcript`, `TranscriptParams`, and `TranscriptOptionalParams`.
   > The confidence threshold for the automatically detected language.
-  > An error will be returned if the langauge confidence is below this threshold.
+  > An error will be returned if the language confidence is below this threshold.
 - Add `language_confidence` to `Transcript`
   > The confidence score for the detected language, between 0.0 (low confidence) and 1.0 (high confidence)
 

package.json

@@ -1,6 +1,6 @@
 {
   "name": "assemblyai",
-  "version": "4.25.1",
+  "version": "4.26.0",
   "description": "The AssemblyAI JavaScript SDK provides an easy-to-use interface for interacting with the AssemblyAI API, which supports async and real-time transcription, as well as the latest LeMUR models.",
   "engines": {
     "node": ">=18"

src/services/streaming/service.ts

@@ -142,14 +142,12 @@ export class StreamingTranscriber {
       );
     }
 
-    if (this.params.speechModel) {
-      if (this.params.speechModel === "u3-pro") {
-        console.warn(
-          "[Deprecation Warning] The speech model `u3-pro` is deprecated and will be removed in a future release. Please use `u3-rt-pro` instead.",
-        );
-      }
-      searchParams.set("speech_model", this.params.speechModel.toString());
+    if (this.params.speechModel === "u3-pro") {
+      console.warn(
+        "[Deprecation Warning] The speech model `u3-pro` is deprecated and will be removed in a future release. Please use `u3-rt-pro` instead.",
+      );
     }
+    searchParams.set("speech_model", this.params.speechModel.toString());
 
     if (this.params.languageDetection !== undefined) {
       searchParams.set(

src/types/streaming/index.ts

@@ -19,7 +19,7 @@ export type StreamingTranscriberParams = {
   keyterms?: string[];
   keytermsPrompt?: string[];
   prompt?: string;
-  speechModel?: StreamingSpeechModel;
+  speechModel: StreamingSpeechModel;
   languageDetection?: boolean;
   inactivityTimeout?: number;
   speakerLabels?: boolean;

tests/unit/streaming.test.ts

@@ -46,6 +46,7 @@ describe("streaming", () => {
       websocketBaseUrl: websocketBaseUrl,
       apiKey: "123",
       sampleRate: 16_000,
+      speechModel: "universal-streaming-english",
     });
     onOpen = jest.fn();
     rt.on("open", onOpen);
@@ -64,12 +65,13 @@ describe("streaming", () => {
     await cleanup();
     WS.clean();
 
-    const wsUrl = `${websocketBaseUrl}?token=123&sample_rate=16000&speaker_labels=true`;
+    const wsUrl = `${websocketBaseUrl}?token=123&sample_rate=16000&speech_model=universal-streaming-english&speaker_labels=true`;
     server = new WS(wsUrl);
     rt = new StreamingTranscriber({
       websocketBaseUrl,
       token: "123",
       sampleRate: 16_000,
+      speechModel: "universal-streaming-english",
       speakerLabels: true,
     });
     onOpen = jest.fn();
@@ -81,12 +83,13 @@ describe("streaming", () => {
     await cleanup();
     WS.clean();
 
-    const wsUrl = `${websocketBaseUrl}?token=123&sample_rate=16000&speaker_labels=true&max_speakers=4`;
+    const wsUrl = `${websocketBaseUrl}?token=123&sample_rate=16000&speech_model=universal-streaming-english&speaker_labels=true&max_speakers=4`;
     server = new WS(wsUrl);
     rt = new StreamingTranscriber({
       websocketBaseUrl,
       token: "123",
       sampleRate: 16_000,
+      speechModel: "universal-streaming-english",
       speakerLabels: true,
       maxSpeakers: 4,
     });
@@ -105,7 +108,7 @@ describe("streaming", () => {
       websocketBaseUrl,
       token: "123",
       sampleRate: 16_000,
-      speechModel: "whisper-rt",
+      speechModel: "whisper-rt" as const,
     });
     onOpen = jest.fn();
     rt.on("open", onOpen);