Open kms9 opened 8 months ago
Bot detected the issue body's language is not English, translate it automatically.
Title: [Feature Request]: Can the voice input function be supported?
Currently you can only chat via text, but voice chat is more convenient
If you call the optional speech recognition interface (openai whipser, domestic voice/or customized interface) and then switch to question and answer call llm, and then switch to voice return through the online tts interface again for playback
No response
No response
Problem Description
目前只能通过文字聊天 但是语音聊天更为方便
Solution Description
如果调用 可选的语音识别接口(openai whipser, 国内一众语音/ 或是自定义的接口) 然后转为问答调用llm, 然后再次通过在线的tts接口转为语音返回 进行播放
Alternatives Considered
No response
Additional Context
No response
Implementing this is more challenging due to the constraints of local storage. Unlike images, which can be stored using base64 in the browser's local storage, this feature presents more difficulties.
我这边有实现,可以看看:https://github.com/vual/ChatGPT-Next-Web-Pro
Bot detected the issue body's language is not English, translate it automatically.
I have an implementation here, you can check it out: https://github.com/vual/ChatGPT-Next-Web-Pro
我这边有实现,可以看看:https://github.com/vual/ChatGPT-Next-Web-Pro
不能直接用 不感兴趣
Bot detected the issue body's language is not English, translate it automatically.
I have an implementation here, you can check it out: https://github.com/vual/ChatGPT-Next-Web-Pro
Cannot be used directly. Not interested.
@kms9 There are plans for this feature, but as @H0llyW00dzZ said, we still expect server-side implementations to support audio files in the future due to storage concerns. In the short term, it may be possible to use the browser API to convert speech directly to text, but it will not be possible to replay the audio.
Problem Description
目前只能通过文字聊天 但是语音聊天更为方便
Solution Description
如果调用 可选的语音识别接口(openai whipser, 国内一众语音/ 或是自定义的接口) 然后转为问答调用llm, 然后再次通过在线的tts接口转为语音返回 进行播放
Alternatives Considered
No response
Additional Context
No response