Gazelle v0.2 is a joint speech-language model that can take in text and audio as input and generate text as output. It is designed for business use cases where latency and conversational quality matter, such as customer support and outbound sales.
1 Impression