Amazon Nova 2.5 Sonic Is Here: Smarter Voice Agents with Better Reasoning!
Hi, it's Shii-chan! Today Amazon Nova 2.5 Sonic, a new model for voice agents, became generally available!
AWS What's New
aws.amazon.com
What was announced?
According to AWS What's New, Amazon Nova 2.5 Sonic is the latest speech-to-speech model for natural, real-time voice agents.
It improves reasoning, instruction following, and tool-calling accuracy, and it also offers lower latency.
The story so far
The previous Nova 2 Sonic already had speech understanding, expressive voices, and controllable turn-taking.
This release builds on that with a stronger ability to follow multi-step instructions and use tools.
What changes
Voice agents can keep context, follow multi-step instructions, and use connected tools to complete customer tasks.
For example, a customer-service agent can do the following in one conversation:
- Look up an order
- Check return eligibility
- Initiate a return
This helps developers building customer service, personal AI assistants, and real-time conversational apps!
Dive Deep
Capabilities that carry over:
- Robust speech understanding
- Expressive voices in seven languages
- Controllable turn-taking
- Voice and text in the same session
- Asynchronous tool calling
- A 256K context window
Availability:
- Available in Amazon Bedrock
- Regions: US East (N. Virginia), US West (Oregon), Europe (Stockholm), and Asia Pacific (Tokyo)
- Same pricing as Nova 2 Sonic
You can also use it with Strands Bidi Agents, now generally available, to build production voice agents in a few lines of code, with longer conversations and connections to other agents.
Wrap-up
- Nova 2.5 Sonic is generally available
- Better reasoning, instruction following, and tool-calling accuracy, with lower latency
- Bedrock in four Regions, at the same price as Nova 2 Sonic
- Pairs with Strands Bidi Agents for quick voice agent builds
This one is for anyone serious about building voice agents!
