LiveVoice AI - Realtime conversational AI
LiveVoice AI Pro and Lite just dropped, allowing you to kickstart your conversational AI projects in minutes!
Utilizing bidirectional web sockets we manage to get extremely low latency (often under 0.2s) while maintaining a very high voice quality with cutting edge models.
We support the full feature stack, like Voice Activation Detection (to interrupt the AI), function calling for all your game logic and context gathering, a ton of additional features!
This all while offering low (or free) generation costs, as this is a single model instead of separate models for each step in the pipeline (SST + TextGen + TTS).
Get it while it is on sale:
LiveVoice AI Pro:
LiveVoice AI Lite:
Full feature list:
Voice Activation Detection (VAD): Interupt AI and only send data when you actually talk
Function Calling: Call functions from the AI model in order to interact with your game, by sending or receiving data
Multilingual interaction: Up to 30 languages supported!
Multi voice: All Gemini voices are supported out of the box!
Platform support: Using Native Unity features to support all native platforms, including Android and VR!
Realtime transcription (of the user AND the AI)
Lightweight API