Amazon plays catchup with new Nova AI models to generate voices and video
1 min read
Summary
Amazon has released a new AI voice model called Nova Sonic, available on its Bedrock developer platform, that handles real-time speech processing and AI voice generation for conversational applications.
Nova Sonic uses one model to recognise speech, convert it to text, generate a response and then convert the text to audio, rather than using separate models for each step.
The company has also upgraded its Nova Reel video model to version 1.1, which can stitch together multiple six-second scenes into a single video of up to two minutes in length, whilst maintaining a consistent style.
Amazon has released no further details about when the technology will be available beyond statements that it is “available to try”.
However, the technology is already being used in Amazon’s new Alexa Plus assistant.