support cartesia tts

lib/synth-audio.js

@@ -3,6 +3,7 @@ const fs = require('fs');
 const bent = require('bent');
 const ttsGoogle = require('@google-cloud/text-to-speech');
 const { PollyClient, SynthesizeSpeechCommand } = require('@aws-sdk/client-polly');
+const { CartesiaClient } = require('@cartesia/cartesia-js');
 
 const sdk = require('microsoft-cognitiveservices-speech-sdk');
 const TextToSpeechV1 = require('ibm-watson/text-to-speech/v1');
@@ -96,7 +97,7 @@ async function synthAudio(client, createHash, retrieveHash, logger, stats, { acc
   logger = logger || noopLogger;
 
   assert.ok(['google', 'aws', 'polly', 'microsoft', 'wellsaid', 'nuance', 'nvidia', 'ibm', 'elevenlabs',
-    'whisper', 'deepgram', 'playht', 'rimelabs', 'verbio'].includes(vendor) ||
+    'whisper', 'deepgram', 'playht', 'rimelabs', 'verbio', 'cartesia'].includes(vendor) ||
   vendor.startsWith('custom'),
   `synthAudio supported vendors are google, aws, microsoft, nuance, nvidia and wellsaid ..etc, not ${vendor}`);
   if ('google' === vendor) {
@@ -157,7 +158,9 @@ async function synthAudio(client, createHash, retrieveHash, logger, stats, { acc
     if (!credentials.deepgram_tts_uri) {
       assert.ok(credentials.api_key, 'synthAudio requires api_key when deepgram is used');
     }
-
+  } else if ('cartesia' === vendor) {
+    assert.ok(credentials.api_key, 'synthAudio requires api_key when cartesia is used');
+    assert.ok(credentials.model_id, 'synthAudio requires model_id when cartesia is used');
   }
 
   const key = makeSynthKey({
@@ -261,6 +264,11 @@ async function synthAudio(client, createHash, retrieveHash, logger, stats, { acc
           credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming, filePath
         });
         break;
+      case 'cartesia':
+        audioBuffer = await synthCartesia(logger, {
+          credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming, filePath
+        });
+        break;
       case 'rimelabs':
         audioBuffer = await synthRimelabs(logger, {
           credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming, filePath
@@ -1067,6 +1075,71 @@ const synthDeepgram = async(logger, {credentials, stats, model, text, renderForC
   }
 };
 
+const synthCartesia = async(logger, {
+  credentials, options, stats, voice, language, text, renderForCaching, disableTtsStreaming
+}) => {
+  const {api_key, model_id, embedding, options: credOpts} = credentials;
+  const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
+
+  if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
+    let params = '';
+    params += `{api_key=${api_key}`;
+    params += `,model_id=${model_id}`;
+    params += ',vendor=cartesia';
+    params += `,voice=${voice}`;
+    params += ',write_cache_file=1';
+    params += `,language=${language}`;
+    params += `,voice_mode=${embedding ? 'embedding' : 'id'}`;
+    if (embedding) params += `,embedding=${embedding}`;
+    if (opts.speed) params += `,speed=${opts.speed}`;
+    if (opts.emotion) params += `,emotion=${opts.emotion}`;
+    params += '}';
+
+    return {
+      filePath: `say:${params}${text.replace(/\n/g, ' ').replace(/\r/g, ' ')}`,
+      servedFromCache: false,
+      rtt: 0
+    };
+  }
+
+  try {
+    const client = new CartesiaClient({ apiKey: api_key });
+    const mp3 = await client.tts.bytes({
+      modelId: model_id,
+      transcript: text,
+      voice: {
+        mode: embedding ? 'embedding' : 'id',
+        ...(embedding ?
+          {
+            embedding: embedding.split(',').map(Number)
+          } :
+          {
+            id: voice
+          }
+        ),
+        ...(opts.speed || opts.emotion && {
+          experimentalControls: {
+            ...(opts.speed && {speed: opts.speed}),
+            ...(opts.emotion && {emotion: opts.emotion}),
+          }
+        })
+      },
+      language: language,
+      outputFormat: {
+        container: 'mp3',
+        bitRate: 128000,
+        sampleRate: 8000
+      },
+    });
+    return Buffer.from(mp3);
+  } catch (err) {
+    logger.info({err}, 'synth Cartesia returned error');
+    stats.increment('tts.count', ['vendor:cartesia', 'accepted:no']);
+    throw err;
+  }
+
+};
+
 const getFileExtFromMime = (mime) => {
   switch (mime) {
     case 'audio/wav':

lib/utils.js

@@ -50,6 +50,7 @@ function getFileExtension({vendor, voice, renderForCaching = false}) {
     case 'elevenlabs':
     case 'rimelabs':
     case 'playht':
+    case 'cartesia':
       if (renderForCaching || JAMBONES_DISABLE_TTS_STREAMING) {
         return mp3Extension;
       } else {