Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.voiceos.com/llms.txt

Use this file to discover all available pages before exploring further.

This guide shows the beta API flow for generating a prompt-aware transcript.

​
1. Use the beta endpoint

Use the VoiceOS beta API endpoint: 
https://beta.api.voiceos.com

​
2. (Optional) Prepare context

Use a chat-style message stack when you want better recognition for domain terms.
If you do not need context, you can skip messages entirely and send only file. 
[
  {
    "role": "system",
    "content": "You are helping inside a TypeScript codebase."
  },
  {
    "role": "user",
    "content": "The relevant files are SessionStreamService.ts and promptableTranscription.ts. The env var is VOICEOS_TEAM_API_KEY."
  }
]

​
3. Send audio

curl https://beta.api.voiceos.com/v1/audio/transcriptions \
  -F "file=@sample.mp3" \
  -F 'messages=[{"role":"system","content":"You are helping inside a TypeScript codebase."},{"role":"user","content":"The relevant files are SessionStreamService.ts and promptableTranscription.ts. The env var is VOICEOS_TEAM_API_KEY."}]' \
  -F 'languages="en"' \
  -F 'response_format=json'

​
4. Read the response

{
  "text": "Please open SessionStreamService.ts and update the VOICEOS_TEAM_API_KEY inside promptableTranscription.ts."
}
Use text as the product-ready transcript output.