voice/gemini_tts

To convert HAL verdict text into audio using Gemini TTS, you pass the HAL text to the Gemini voice, which vocalizes the exact ledger content verbatim without rephrasing, generating an audio artifact for real-time or archival use.

Audio provenance in text-to-speech ledger outputs is a document pairing the generated audio with the original HAL text and receipt, detailing the model, voice, seeds, and cryptographic hashes to ensure verifiable vocal delivery.

Yes, you need a Gemini API key to access the text-to-speech service, alongside passing K8 non-determinism boundary validation to generate the audio artifact and its accompanying provenance file.

Yes, you can use Gemini TTS to vocalize HAL JSON streams for real-time voice interfaces, speaking the HAL output exactly as written in the ledger for live verifiable vocal delivery.

No, Gemini TTS does not rephrase the original text when converting HAL verdicts to speech; it speaks the HAL JSON stream exactly as written in the ledger to preserve verbatim content.

The best way to archive HELEN decisions as auditable audio records is to convert the HAL verdict text using Gemini TTS and generate a provenance document pairing the audio with the HAL text and cryptographic receipt.