← kwindla hultman kramer

New speech-to-speech model from Amazon!

May 9, 2025

New speech-to-speech model from Amazon!

Nova Sonic has a bidirectional streaming API, multiple voices, tool calling, and good performance on standard benchmarks.

There's full support for Nova Sonic in the most recent @pipecat_ai release (0.0.67).

Link to docs and to Pipecat…

Amazon Nova Sonic overview page.

[1]

Example code. (This is what I used to record the video above.)

[2]

  1. https://aws.amazon.com/ai/generative-ai/nova/speech/
  2. https://github.com/pipecat-ai/pipecat/blob/main/examples/foundational/39-aws-nova-sonic.py