Use existing sse streamer to stream tts audio before sse client is removed

packages/server/src/utils/SSEStreamer.ts

@@ -268,4 +268,30 @@ export class SSEStreamer implements IServerSideEventStreamer {
             client.response.write('message:\ndata:' + JSON.stringify(clientResponse) + '\n\n')
         }
     }
+
+    streamTTSDataEvent(chatId: string, audioChunk: string): void {
+        const client = this.clients[chatId]
+        console.log('clients', this.clients)
+        console.log('client', client)
+        if (client) {
+            const clientResponse = {
+                event: 'tts_data',
+                data: audioChunk
+            }
+            client.response.write('message:\ndata:' + JSON.stringify(clientResponse) + '\n\n')
+        }
+    }
+
+    streamTTSEndEvent(chatId: string): void {
+        const client = this.clients[chatId]
+        console.log('clients', this.clients)
+        console.log('client', client)
+        if (client) {
+            const clientResponse = {
+                event: 'tts_end',
+                data: {}
+            }
+            client.response.write('message:\ndata:' + JSON.stringify(clientResponse) + '\n\n')
+        }
+    }
 }

packages/server/src/utils/buildAgentflow.ts

@@ -11,8 +11,7 @@ import {
     IMessage,
     IServerSideEventStreamer,
     convertChatHistoryToText,
-    generateFollowUpPrompts,
+    generateFollowUpPrompts
-    convertTextToSpeech
 } from 'flowise-components'
 import {
     IncomingAgentflowInput,
@@ -58,6 +57,7 @@ import { ChatMessage } from '../database/entities/ChatMessage'
 import { Telemetry } from './telemetry'
 import { getWorkspaceSearchOptions } from '../enterprise/utils/ControllerServiceUtils'
 import { UsageCacheManager } from '../UsageCacheManager'
+import { generateTTSForResponseStream, shouldAutoPlayTTS } from './buildChatflow'
 
 interface IWaitingNode {
     nodeId: string
@@ -136,59 +136,6 @@ interface IExecuteNodeParams {
     subscriptionId: string
 }
 
-// Helper function to check if auto-play TTS is enabled
-const shouldAutoPlayTTS = (textToSpeechConfig: string | undefined | null): boolean => {
-    if (!textToSpeechConfig) return false
-    try {
-        const config = typeof textToSpeechConfig === 'string' ? JSON.parse(textToSpeechConfig) : textToSpeechConfig
-        // Check each provider to see if any has autoPlay enabled and status true
-        for (const providerKey in config) {
-            const provider = config[providerKey]
-            if (provider && provider.status === true && provider.autoPlay === true) {
-                return true
-            }
-        }
-        return false
-    } catch (error) {
-        return false
-    }
-}
-
-// Helper function to generate TTS for response
-const generateTTSForResponse = async (
-    responseText: string,
-    textToSpeechConfig: string | undefined,
-    options: ICommonObject
-): Promise<Buffer | null> => {
-    try {
-        if (!textToSpeechConfig) return null
-        const config = typeof textToSpeechConfig === 'string' ? JSON.parse(textToSpeechConfig) : textToSpeechConfig
-
-        // Find the active provider configuration
-        let activeProviderConfig = null
-        for (const providerKey in config) {
-            const provider = config[providerKey]
-            if (provider && provider.status === true) {
-                activeProviderConfig = {
-                    name: providerKey,
-                    credentialId: provider.credentialId,
-                    voice: provider.voice,
-                    model: provider.model
-                }
-                break
-            }
-        }
-
-        if (!activeProviderConfig) return null
-
-        const audioBuffer = await convertTextToSpeech(responseText, activeProviderConfig, options)
-        return audioBuffer
-    } catch (error) {
-        logger.error(`[server]: TTS generation failed: ${getErrorMessage(error)}`)
-        return null
-    }
-}
-
 interface IExecuteAgentFlowParams extends Omit<IExecuteFlowParams, 'incomingInput'> {
     incomingInput: IncomingAgentflowInput
 }
@@ -2092,7 +2039,6 @@ export const executeAgentFlow = async ({
 
     if (sessionId) result.sessionId = sessionId
 
-    /*** Auto-play TTS Logic ***/
     if (shouldAutoPlayTTS(chatflow.textToSpeech) && result.text) {
         const options = {
             orgId,
@@ -2102,14 +2048,8 @@ export const executeAgentFlow = async ({
             databaseEntities
         }
 
-        const audioBuffer = await generateTTSForResponse(result.text, chatflow.textToSpeech, options)
+        if (sseStreamer) {
-        if (audioBuffer) {
+            await generateTTSForResponseStream(result.text, chatflow.textToSpeech, options, chatId, sseStreamer)
-            const audioBase64 = audioBuffer.toString('base64')
-
-            // Agent flows are always streamed, so send audio via SSE
-            if (sseStreamer) {
-                sseStreamer.streamAudioEvent(chatId, audioBase64)
-            }
         }
     }
 

packages/server/src/utils/buildChatflow.ts

@@ -7,6 +7,7 @@ import {
     IFileUpload,
     convertSpeechToText,
     convertTextToSpeech,
+    convertTextToSpeechStream,
     ICommonObject,
     addSingleFileToStorage,
     generateFollowUpPrompts,
@@ -17,7 +18,8 @@ import {
     getFileFromUpload,
     removeSpecificFileFromUpload,
     EvaluationRunner,
-    handleEscapeCharacters
+    handleEscapeCharacters,
+    IServerSideEventStreamer
 } from 'flowise-components'
 import { StatusCodes } from 'http-status-codes'
 import {
@@ -71,12 +73,10 @@ import { executeAgentFlow } from './buildAgentflow'
 import { Workspace } from '../enterprise/database/entities/workspace.entity'
 import { Organization } from '../enterprise/database/entities/organization.entity'
 
-// Helper function to check if auto-play TTS is enabled
 const shouldAutoPlayTTS = (textToSpeechConfig: string | undefined | null): boolean => {
     if (!textToSpeechConfig) return false
     try {
         const config = typeof textToSpeechConfig === 'string' ? JSON.parse(textToSpeechConfig) : textToSpeechConfig
-        // Check each provider to see if any has autoPlay enabled and status true
         for (const providerKey in config) {
             const provider = config[providerKey]
             if (provider && provider.status === true && provider.autoPlay === true) {
@@ -85,21 +85,22 @@ const shouldAutoPlayTTS = (textToSpeechConfig: string | undefined | null): boole
         }
         return false
     } catch (error) {
+        logger.error(`Error parsing textToSpeechConfig: ${getErrorMessage(error)}`)
         return false
     }
 }
 
-// Helper function to generate TTS for response
+const generateTTSForResponseStream = async (
-const generateTTSForResponse = async (
     responseText: string,
     textToSpeechConfig: string | undefined,
-    options: ICommonObject
+    options: ICommonObject,
-): Promise<Buffer | null> => {
+    chatId: string,
+    sseStreamer: IServerSideEventStreamer
+): Promise<void> => {
     try {
-        if (!textToSpeechConfig) return null
+        if (!textToSpeechConfig) return
         const config = typeof textToSpeechConfig === 'string' ? JSON.parse(textToSpeechConfig) : textToSpeechConfig
 
-        // Find the active provider configuration
         let activeProviderConfig = null
         for (const providerKey in config) {
             const provider = config[providerKey]
@@ -114,13 +115,24 @@ const generateTTSForResponse = async (
             }
         }
 
-        if (!activeProviderConfig) return null
+        if (!activeProviderConfig) return
 
-        const audioBuffer = await convertTextToSpeech(responseText, activeProviderConfig, options)
+        await convertTextToSpeechStream(
-        return audioBuffer
+            responseText,
+            activeProviderConfig,
+            options,
+            (chunk: Buffer) => {
+                const audioBase64 = chunk.toString('base64')
+                logger.info(`Received TTS chunk: ${audioBase64}`)
+                sseStreamer.streamTTSDataEvent(chatId, audioBase64)
+            },
+            () => {
+                sseStreamer.streamTTSEndEvent(chatId)
+            }
+        )
     } catch (error) {
-        logger.error(`[server]: TTS generation failed: ${getErrorMessage(error)}`)
+        logger.error(`[server]: TTS streaming failed: ${getErrorMessage(error)}`)
-        return null
+        sseStreamer.streamTTSEndEvent(chatId)
     }
 }
 
@@ -880,8 +892,6 @@ export const executeFlow = async ({
         if (Object.keys(setVariableNodesOutput).length) result.flowVariables = setVariableNodesOutput
 
         if (shouldAutoPlayTTS(chatflow.textToSpeech) && result.text) {
-            logger.info('[server]: Generating TTS for response')
-            logger.info(`[server/executeFlow]: TTS config: ${JSON.stringify(chatflow.textToSpeech)}`)
             const options = {
                 orgId,
                 chatflowid,
@@ -890,15 +900,10 @@ export const executeFlow = async ({
                 databaseEntities
             }
 
-            const audioBuffer = await generateTTSForResponse(result.text, chatflow.textToSpeech, options)
+            if (streaming && sseStreamer) {
-            if (audioBuffer) {
+                await generateTTSForResponseStream(result.text, chatflow.textToSpeech, options, chatId, sseStreamer)
-                const audioBase64 = audioBuffer.toString('base64')
+            } else if (sseStreamer) {
-
+                await generateTTSForResponseStream(result.text, chatflow.textToSpeech, options, chatId, sseStreamer)
-                if (streaming && sseStreamer) {
-                    sseStreamer.streamAudioEvent(chatId, audioBase64)
-                } else {
-                    result.audioData = audioBase64
-                }
             }
         }
 
@@ -1129,3 +1134,5 @@ const incrementFailedMetricCounter = (metricsProvider: IMetricsProvider, isInter
         )
     }
 }
+
+export { shouldAutoPlayTTS, generateTTSForResponseStream }