Use API Keys, OpenAI-Compatible STT, and Shortcuts

Use the following standard OpenAI Audio parameters across all compatible clients:

All tools below use standard OpenAI Audio Transcriptions and can be wired to Vowise in under a minute:

For **Orca**, create or edit `~/.orca/speech-config.json` on the machine running the Orca daemon. Both the desktop composer and the mobile web composer will route voice recordings through Vowise:

{
  "apiKey": "vowise_YOUR_API_KEY",
  "baseUrl": "https://api.vowise.com/v1",
  "model": "vowise-pro",
  "language": "zh"
}

For **Paseo**, configure `providers.openai` and `features.dictation` inside `~/.paseo/config.json`:

{
  "providers": {
    "openai": {
      "apiKey": "vowise_YOUR_API_KEY",
      "baseUrl": "https://api.vowise.com/v1"
    }
  },
  "features": {
    "dictation": {
      "enabled": true,
      "stt": {
        "provider": "openai",
        "model": "vowise-pro",
        "language": "zh"
      }
    }
  }
}

Test from the terminal with `curl` or use the official OpenAI Python/Node SDKs without installing any custom client:

curl -X POST "https://api.vowise.com/v1/audio/transcriptions" \
  -H "Authorization: Bearer vowise_YOUR_API_KEY" \
  -F "[email protected]" \
  -F "model=vowise-pro" \
  -F "language=zh"

from openai import OpenAI

client = OpenAI( base_url="https://api.vowise.com/v1", api_key="vowise_YOUR_API_KEY", )

with open("sample.wav", "rb") as audio: tx = client.audio.transcriptions.create( model="vowise-pro", file=audio, language="zh", ) print(tx.text) ```

Docs can teach the request shape, but examples should use placeholders instead of real secrets, reusable passwords, or private URLs.