Hi Google Cloud Team,
I am requesting to have my project allowlisted for Gemini native audio output (response_modalities: ["AUDIO"]) and the Multimodal Live API.
My project id: 295327889536
I am developing an AI-powered semi-structured interviewer designed specifically for development program evaluations.
I am migrating my architecture from a traditional STT → LLM → TTS pipeline to a native audio-to-audio multimodal flow using gemini-2.5-flash and gemini-2.5-pro. This architectural shift is critical for my use case to reduce latency and natively handle multi-language switching (e.g., English, Vietnamese, Thai, French, Spanish) during live, conversational evaluations without relying on transcription steps.
I understand the safety considerations surrounding raw audio generation. This application is deployed in a controlled, secure environment strictly for research, monitoring, and evaluation purposes, ensuring all interactions are consensual and safely managed.
Could you please advise on the next steps or any additional information required to allowlist this project for native audio output?
Thank you,
Ali