Voice Enhancement

The voice enhancement service uses AI to clean up and improve voice transcripts.

Enhance transcript

POST /api/voice/enhance

Enhance a voice transcript using the configured LLM provider.

Field	Type	Required	Description
`transcript`	string	Yes	Raw transcript text
`language`	string	No	Language code (default: zh-CN)
`options.enableSelfCorrection`	boolean	No	Enable auto-correction
`options.enableListFormatting`	boolean	No	Format lists
`options.enableFillerRemoval`	boolean	No	Remove filler words
`options.enableToneAdjustment`	boolean	No	Adjust tone
`options.targetTone`	string	No	Target tone: formal/casual/professional

const result = await client.enhanceVoice({
  transcript: 'Um, so I think we should, like, deploy tomorrow maybe...',
  language: 'en-US',
  options: {
    enableSelfCorrection: true,
    enableFillerRemoval: true,
  },
})

Enhance transcript (query params)

GET /api/voice/enhance?transcript=...&language=en-US

Same as the POST version, but uses query parameters.

Param	Type	Required	Description
`transcript`	string	Yes	Raw transcript text
`language`	string	No	Language code
`enableSelfCorrection`	boolean	No	Auto-correction
`enableListFormatting`	boolean	No	List formatting
`enableFillerRemoval`	boolean	No	Filler word removal
`enableToneAdjustment`	boolean	No	Tone adjustment
`targetTone`	string	No	formal/casual/professional

const result = await client.enhanceVoiceQuery({
  transcript: 'Hello world',
  language: 'en-US',
  enableFillerRemoval: true,
})

Get voice config

GET /api/voice/config

Returns the current voice enhancement configuration (API key redacted).

const config = await client.getVoiceConfig()

Update voice config

POST /api/voice/config

Admin only. Update the LLM configuration for voice enhancement.

Field	Type	Required	Description
`provider`	string	Yes	openai / anthropic / alibaba / custom
`apiKey`	string	Yes	Provider API key
`baseUrl`	string	No	Custom base URL
`model`	string	No	Model name
`temperature`	number	No	0-2
`maxTokens`	number	No	Max response tokens
`timeout`	number	No	Request timeout (ms)
`enabled`	boolean	No	Enable/disable service

await client.updateVoiceConfig({
  provider: 'openai',
  apiKey: 'sk-...',
  model: 'gpt-4',
  enabled: true,
})

Health check

GET /api/voice/health

Admin only. Checks if the voice enhancement service is configured and operational.

const health = await client.voiceHealthCheck()

ON THIS PAGE

#Voice Enhancement

#Enhance transcript

#Enhance transcript (query params)

#Get voice config