AWS Puts PolyAI Dialog-RSN-1 Into SageMaker JumpStart at $45 an Hour
Dialog-RSN-1 enters SageMaker JumpStart with one-click deployment and four optimization profiles. Inference costs $45-$110 per host hour, and billing runs even when no client is connected.
By Daniel Okafor
4 min read

What's News
- AWS added PolyAI Dialog-RSN-1 to the SageMaker JumpStart catalog on November 19, 2026.
- Real-time inference costs $45 per host hour on ml.p4d.24xlarge and $110 on ml.p5.48xlarge, billed until deletion even when idle.
- Median first-token response is below 300 milliseconds; a typical full caller exchange runs about 390 milliseconds in production.
- The model returns only text or function calls, so every deployment must add its own text-to-speech layer.
- The context window is 32,000 tokens, the model is English-only, and connections cap at 30 minutes.
PolyAI's Dialog-RSN-1 joined the Amazon SageMaker JumpStart catalog on November 19, 2026, bringing one-click deployment, evaluation tools and four optimization profiles to the audio-native customer-service model. Real-time inference starts at $45 per host hour on ml.p4d.24xlarge and rises to $110 per host hour on ml.p5.48xlarge. The endpoint runs inside a customer's own AWS account.
The catalog listing is the third step in a sequence. PolyAI introduced Dialog-RSN-1 on July 30, 2026 and announced SageMaker endpoint deployment through AWS Marketplace on October 6, 2026. JumpStart adds discovery, evaluation and preconfigured deployment on top of the Marketplace subscription customers still complete.
What does JumpStart actually change?
From the SageMaker AI console's Model Hub, a customer selects Dialog-RSN-1 and chooses Deploy. JumpStart creates a real-time inference endpoint on a preconfigured serving stack. That removes manual container selection and the need to benchmark dozens of infrastructure combinations by hand. Teams can deploy with a few lines of SageMaker Python SDK code and fold the process into their CI/CD pipelines.
Before deployment, the model card displays P50 latency, time to first token and throughput. The four profiles split by priority:
- Latency minimizes time to first token.
- Throughput maximizes tokens per second.
- Cost minimizes infrastructure cost per request.
- Balanced serves general-purpose use.
Those choices tie the serving configuration to call volume, responsiveness and spending requirements instead of applying one default to every workload.
Evaluation sits in the same catalog entry. PolyAI's sample notebook streams recorded caller audio and prints the model's turn decisions, replies and transcripts. Teams can review the overview and samples before deployment; streaming their own audio through the model requires a live endpoint.
How does the model handle a live call?
Dialog-RSN-1 processes caller audio directly and combines turn-taking, speech recognition, response generation and function calling in one model. PolyAI says direct audio preserves tone, hesitation and timing that a cascaded system can lose when speech recognition reduces a call to a transcript.
The mechanics are specific. A voice-activity detector triggers each probe, and the model's first output token records EMPTY, ONGOING or COMPLETE, placing the final turn-taking decision with the model itself.
The output path is separate. Dialog-RSN-1 returns streamed reply text or a function call, with the caller transcript arriving separately. It does not synthesize speech. Every deployment must supply its own text-to-speech layer, preserving the customer's choice of voice while adding a component outside the model endpoint.
PolyAI's Marketplace listing reports a median first-token response below 300 milliseconds and about 390 milliseconds for a typical full caller exchange in production. The gap matters: model latency alone does not describe the delay in a complete voice deployment that includes speech synthesis, telephony and external tools.
What are the integration limits and the billing trap?
The endpoint resides in the customer's virtual private cloud, uses its AWS Identity and Access Management roles and is monitored through Amazon CloudWatch. Because inference runs within that account, call audio and transcripts remain within infrastructure the customer controls — a design relevant to data-residency and audit requirements.
The endpoint implements the OpenAI Realtime protocol. PolyAI's client bridge exposes a local WebSocket for Realtime clients and carries their events over SageMaker's signed bidirectional HTTP/2 stream; applications can also invoke that stream directly. Customers retain responsibility for telephony, orchestration, business logic and tool execution, and PolyAI's integration guidance calls for validating dates, reference numbers and amounts before acting on tool calls.
The constraints are firm:
- Audio input is limited to base64-encoded PCM16 mono at 16 or 24 kHz; no G.711, Opus or stereo support.
- The context window is 32,000 tokens.
- The model is English-only.
- Connections have a 30-minute hard cap and a five-minute idle timeout.
- State does not carry across reconnects, so applications must restore relevant context when establishing a new session.
Billing carries the sharpest operational risk. A host hour is one hour of runtime for the compute instance, additional AWS infrastructure charges may apply, and billing starts when the endpoint enters service and continues until deletion — even when no client is connected. AWS's deployment examples delete the endpoint, endpoint configuration and model during cleanup.
JumpStart's main contribution is a shorter operational path from evaluation to a governed AWS endpoint. The remaining production work sits around the model: telephony, speech synthesis, tool validation, session recovery and control of an endpoint that accrues charges while idle. The business case therefore depends on call-handling results from a customer's own audio, measured against those continuing engineering and hosting costs.
Source: unite.ai
More from Daniel Okafor
Show full bio
Correspondent covering money, funding and market shifts around entrepreneurship, startups and venture funding at Business Bearings.
2,420 articles