Dialog-RSN-1, the fastest reasoning model for real-time, complex dialog, is available today on AWS SageMaker. It's the model at the core of frontier dialog: agents that hear a live call, reason about it, and act in real time.
As an AWS customer, you can subscribe to Dialog-RSN-1 through AWS Marketplace and deploy it as a SageMaker endpoint inside your own AWS environment. That means the model best suited to handle your customer conversations is now just as easy to stand up as any other SageMaker workload.
Deploying Dialog-RSN-1 inside AWS
Dialog-RSN-1 is the first PolyAI model available through AWS Marketplace, a result of our active co-sell relationship as an AWS ISV partner. It runs inside your own AWS environment on the same AWS cloud that powers PolyAI's compute and model training and serving.
All in a single pass, Dialog-RSN-1 perceives a caller's audio directly, reasons about what they need, and responds under 300 milliseconds. On SageMaker, it deploys as an endpoint like any other model you run there: reachable from your own architecture, billed against AWS spend you've already committed, and covered by the same VPC, IAM roles, and monitoring you already have in place.
You keep ownership of orchestration, telephony integration, and business logic. Dialog-RSN-1 takes off your plate the hardest part to build yourself: understanding a live call and responding to it in real time.
To get started, click here to find Dialog-RSN-1 on AWS Marketplace, then deploy it straight into a SageMaker endpoint in your own account.
Built for maximum speed and accuracy
Most models built for live calls make you choose between speed and understanding. The fast ones keep pace with the conversation but miss much of the meaning, and the ones that reason carefully leave callers waiting in silence. Dialog-RSN-1 deletes the tradeoff.
For the enterprises that build on PolyAI, that means dialog agents that hear and handle calls the way their best human agents would. They catch hesitation, hear frustration before it escalates, and know exactly when to speak.
In fact, Dialog-RSN-1 scores better than frontier models on the benchmarks that matter most for real-time customer conversations. That's the job it was built for, and the results are already showing up in production, from an 11% lift in call containment for a restaurant group to a 37% cut in latency for an insurance company.
A model that hears calls the way humans do
Dialog-RSN-1 is a single model built for real-time conversation, not a chain of them frankensteined together to cope. There's no separate speech-to-text step running behind the scenes, no separate model guessing when to talk, and no handoff along the way where tone or context gets lost.
In practice, that means your customers get what they need faster, without repeating themselves:
- Barge-in keeps working even when call volume spikes, instead of getting switched off to protect performance.
- Response time stays under 300 milliseconds, so the model isn't lagging behind the conversation while it reasons through what to say.
- And transcription comes out stronger, because Dialog-RSN-1 is reading the actual audio, not someone else's best guess at it.
Altogether, Dialog-RSN-1 makes the difference between a caller who has to struggle through a conversation and one who gets what they need, when they need it.
Get started today with Dialog-RSN-1 on AWS Marketplace.
Learn more about Dialog-RSN-1
For more detail on the architecture behind our frontier dialog model and the evaluation results behind our benchmarks, read our Dialog-RSN-1 launch post.