# Low-latency dialogue

The world's fastest and most efficient edge-deployable audio language model, a 2.6B parameter multimodal model capable of processing both text and audio inputs.