Summary

Add support for Google Cloud Speech-to-Text and Text-to-Speech APIs as providers for audio transcription and TTS.

Motivation

• Users with GCP accounts already have these APIs enabled and want to use their $300 free trial credits
• GCP offers generous free tiers:
  • Speech-to-Text: 60 minutes/month free
  • Text-to-Speech: 4 million characters/month free
• GCP Speech-to-Text supports 125+ languages with excellent quality
• GCP Text-to-Speech offers 220+ voices in 40+ languages
• Authentication can reuse existing ADC (Application Default Credentials) from gcloud auth application-default login

Proposed Implementation

Speech-to-Text (Audio Transcription)

Add google-cloud as a provider option in tools.media.audio:

{
  "tools": {
    "media": {
      "audio": {
        "models": [
          {
            "provider": "google-cloud",
            "model": "latest_long",
            "capabilities": ["audio"],
            "language": "zh-CN"
          }
        ]
      }
    }
  }
}

Text-to-Speech

Add google-cloud as a provider option in messages.tts:

{
  "messages": {
    "tts": {
      "provider": "google-cloud",
      "googleCloud": {
        "voice": "cmn-CN-Wavenet-A",
        "languageCode": "cmn-CN"
      }
    }
  }
}

Authentication

Should support:

1. ADC (Application Default Credentials) - already used by many GCP users
2. Service account JSON key file (optional)

Related

This complements #9729 (Vertex AI Gemini support) - together they would allow full utilization of GCP free credits within OpenClaw.

Additional Context

• GCP Speech API: https://cloud.google.com/speech-to-text
• GCP TTS API: https://cloud.google.com/text-to-speech
• Node.js clients available: @google-cloud/speech, @google-cloud/text-to-speech

Thanks for considering!