Hello Google team,
I’d like to request allowlist / preview access for the experimental Lyria RealTime model on my Gemini API project.
Project ID: gen-lang-client-02494047..
Access path: Gemini API (google-genai SDK / Live Music WebSocket)
Model requested: models/lyria-realtime-exp
Billing: enabled
Key type: AI Studio auth key
Use case:
I am building a Unity-based research prototype titled:
“Confidence-Aware Personalized Adaptive Audio Controller for AI-Generated VR Meditation”.
The system uses:
- personalized audio preferences
- simulated stress and confidence inputs
- safe adaptive audio control in Unity
- instrumental meditation / ambient music only
The role of Lyria RealTime in this prototype is:
- generate and steer instrumental meditation music in real time
- allow smooth prompt and parameter updates during a session
- support safe adaptive control for VR relaxation research
Expected usage:
- research prototype
- low concurrency
- single-user / small-scale testing first
- no copyrighted reference audio uploads
- no voice cloning
- no public large-scale deployment at this stage
Current symptom:
The same project/key can use non-realtime Lyria clip generation, but attempts to open a Lyria RealTime session for models/lyria-realtime-exp fail during session open. My local backend and websocket bridge are functioning correctly, so this appears to be project access / allowlist related.
If needed, I can provide:
- exact project ID
- sample request code
- backend logs
- request/response details
Thank you.