diff --git a/api/core/model_runtime/model_providers/openai/tts/tts.py b/api/core/model_runtime/model_providers/openai/tts/tts.py
index 608ed897e096f..d3fcf731f15ac 100644
--- a/api/core/model_runtime/model_providers/openai/tts/tts.py
+++ b/api/core/model_runtime/model_providers/openai/tts/tts.py
@@ -114,7 +114,8 @@ def _tts_invoke_streaming(self, model: str, credentials: dict, content_text: str
# doc: https://platform.openai.com/docs/guides/text-to-speech
credentials_kwargs = self._to_credential_kwargs(credentials)
client = OpenAI(**credentials_kwargs)
- if not voice or voice not in self.get_tts_model_voices(model=model, credentials=credentials):
+ model_support_voice = [x.get("value") for x in self.get_tts_model_voices(model=model, credentials=credentials)]
+ if not voice or voice not in model_support_voice:
voice = self._get_model_default_voice(model, credentials)
word_limit = self._get_model_word_limit(model, credentials)
if len(content_text) > word_limit:
diff --git a/web/app/components/app/configuration/config-voice/param-config-content.tsx b/web/app/components/app/configuration/config-voice/param-config-content.tsx
index cced3b045849c..9b0d5bbb69e23 100644
--- a/web/app/components/app/configuration/config-voice/param-config-content.tsx
+++ b/web/app/components/app/configuration/config-voice/param-config-content.tsx
@@ -31,12 +31,12 @@ const VoiceParamConfig: FC = () => {
let languageItem = languages.find(item => item.value === textToSpeechConfig.language)
const localLanguagePlaceholder = languageItem?.name || t('common.placeholder.select')
- if (languages && !languageItem)
+ if (languages && !languageItem && languages.length > 0)
languageItem = languages[0]
const language = languageItem?.value
const voiceItems = useSWR({ appId, language }, fetchAppVoices).data
let voiceItem = voiceItems?.find(item => item.value === textToSpeechConfig.voice)
- if (voiceItems && !voiceItem)
+ if (voiceItems && !voiceItem && voiceItems.length > 0)
voiceItem = voiceItems[0]
const localVoicePlaceholder = voiceItem?.name || t('common.placeholder.select')
@@ -125,9 +125,11 @@ const VoiceParamConfig: FC = () => {
{t('appDebug.voice.voiceSettings.voice')}
{
+ if (!value.value)
+ return
setTextToSpeechConfig({
...textToSpeechConfig,
voice: String(value.value),
diff --git a/web/app/components/app/configuration/features/chat-group/text-to-speech/index.tsx b/web/app/components/app/configuration/features/chat-group/text-to-speech/index.tsx
index 4c5db2251312b..72d617c3c371c 100644
--- a/web/app/components/app/configuration/features/chat-group/text-to-speech/index.tsx
+++ b/web/app/components/app/configuration/features/chat-group/text-to-speech/index.tsx
@@ -41,6 +41,7 @@ const TextToSpeech: FC = () => {
)}
diff --git a/web/app/components/app/text-generate/item/index.tsx b/web/app/components/app/text-generate/item/index.tsx
index 3313c987c9343..9794967d9d1f6 100644
--- a/web/app/components/app/text-generate/item/index.tsx
+++ b/web/app/components/app/text-generate/item/index.tsx
@@ -29,6 +29,7 @@ import { useStore as useAppStore } from '@/app/components/app/store'
import WorkflowProcessItem from '@/app/components/base/chat/chat/answer/workflow-process'
import type { WorkflowProcess } from '@/app/components/base/chat/types'
import type { SiteInfo } from '@/models/share'
+import { useChatContext } from '@/app/components/base/chat/chat/context'
const MAX_DEPTH = 3
@@ -127,6 +128,10 @@ const GenerationItem: FC = ({
const [childFeedback, setChildFeedback] = useState({
rating: null,
})
+ const {
+ config,
+ } = useChatContext()
+
const setCurrentLogItem = useAppStore(s => s.setCurrentLogItem)
const setShowPromptLogModal = useAppStore(s => s.setShowPromptLogModal)
@@ -430,6 +435,7 @@ const GenerationItem: FC = ({
>
)}
diff --git a/web/app/components/base/audio-btn/audio.player.manager.ts b/web/app/components/base/audio-btn/audio.player.manager.ts
index 03e9e21f93f6a..17d92f8dc25f7 100644
--- a/web/app/components/base/audio-btn/audio.player.manager.ts
+++ b/web/app/components/base/audio-btn/audio.player.manager.ts
@@ -41,7 +41,7 @@ export class AudioPlayerManager {
}
this.msgId = id
- this.audioPlayers = new AudioPlayer(url, isPublic, id, msgContent, callback)
+ this.audioPlayers = new AudioPlayer(url, isPublic, id, msgContent, voice, callback)
return this.audioPlayers
}
}
diff --git a/web/app/components/base/audio-btn/audio.ts b/web/app/components/base/audio-btn/audio.ts
index 638626bf8ac69..a61fd085d44c4 100644
--- a/web/app/components/base/audio-btn/audio.ts
+++ b/web/app/components/base/audio-btn/audio.ts
@@ -23,12 +23,13 @@ export default class AudioPlayer {
isPublic: boolean
callback: ((event: string) => {}) | null
- constructor(streamUrl: string, isPublic: boolean, msgId: string | undefined, msgContent: string | null | undefined, callback: ((event: string) => {}) | null) {
+ constructor(streamUrl: string, isPublic: boolean, msgId: string | undefined, msgContent: string | null | undefined, voice: string | undefined, callback: ((event: string) => {}) | null) {
this.audioContext = new AudioContext()
this.msgId = msgId
this.msgContent = msgContent
this.url = streamUrl
this.isPublic = isPublic
+ this.voice = voice
this.callback = callback
// Compatible with iphone ios17 ManagedMediaSource
@@ -154,7 +155,6 @@ export default class AudioPlayer {
this.mediaSource?.endOfStream()
clearInterval(endTimer)
}
- console.log('finishStream endOfStream endTimer')
}, 10)
}
@@ -169,7 +169,6 @@ export default class AudioPlayer {
const arrayBuffer = this.cacheBuffers.shift()!
this.sourceBuffer?.appendBuffer(arrayBuffer)
}
- console.log('finishStream timer')
}, 10)
}
diff --git a/web/app/components/base/audio-btn/index.tsx b/web/app/components/base/audio-btn/index.tsx
index 48081c170c6d2..675f58b53039c 100644
--- a/web/app/components/base/audio-btn/index.tsx
+++ b/web/app/components/base/audio-btn/index.tsx
@@ -65,11 +65,11 @@ const AudioBtn = ({
}
const handleToggle = async () => {
if (audioState === 'playing' || audioState === 'loading') {
- setAudioState('paused')
+ setTimeout(() => setAudioState('paused'), 1)
AudioPlayerManager.getInstance().getAudioPlayer(url, isPublic, id, value, voice, audio_finished_call).pauseAudio()
}
else {
- setAudioState('loading')
+ setTimeout(() => setAudioState('loading'), 1)
AudioPlayerManager.getInstance().getAudioPlayer(url, isPublic, id, value, voice, audio_finished_call).playAudio()
}
}
diff --git a/web/app/components/base/chat/chat/answer/operation.tsx b/web/app/components/base/chat/chat/answer/operation.tsx
index d46aa3437596b..3d52477d0bbe4 100644
--- a/web/app/components/base/chat/chat/answer/operation.tsx
+++ b/web/app/components/base/chat/chat/answer/operation.tsx
@@ -125,6 +125,7 @@ const Operation: FC = ({
id={id}
value={content}
noCache={false}
+ voice={config?.text_to_speech?.voice}
className='hidden group-hover:block'
/>
>
diff --git a/web/app/components/base/features/feature-panel/text-to-speech/param-config-content.tsx b/web/app/components/base/features/feature-panel/text-to-speech/param-config-content.tsx
index ea1d789d0a129..a5a2eb7bb70e3 100644
--- a/web/app/components/base/features/feature-panel/text-to-speech/param-config-content.tsx
+++ b/web/app/components/base/features/feature-panel/text-to-speech/param-config-content.tsx
@@ -149,7 +149,7 @@ const VoiceParamConfig = ({
{t('appDebug.voice.voiceSettings.voice')}
{
handleChange({