Sold by: Salient Engineering
Deployed on AWS
Free Trial
GPU-accelerated multilingual speech-to-text supporting 40 language locales with real-time streaming and batch transcription, deployed entirely within your AWS VPC.
Overview
Nemotron 3.5 ASR - Enterprise Multilingual Speech-to-Text
Nemotron 3.5 ASR is a production-ready, GPU-accelerated speech-to-text appliance powered by NVIDIA, designed to help organizations add accurate voice transcription to contact centers, AI assistants, meeting platforms, media workflows, and more. Because it runs entirely within your AWS VPC as an AMI, all audio data stays on your infrastructure - no data leaves the instance, providing a strong data-sovereignty posture for privacy-sensitive workloads.
Key Capabilities
- Real-time streaming and batch transcription from a single deployment
- 40 language-locales with optional automatic language detection
- Built-in punctuation and capitalization for clean, readable transcripts
- Fine-tuning support to improve recognition for specialized vocabularies and industry terminology
- Configurable performance settings to balance latency and accuracy per workload
Deployment Details
- Supported AWS instance types: GPU-accelerated instances (e.g., g4dn.xlarge, g4dn.2xlarge, g4dn.4xlarge, g4dn.8xlarge) with sufficient GPU memory for model inference
- Operating system: Ubuntu-based AMI 26.04 with NVIDIA GPU drivers pre-installed
- Networking: Configure security groups to allow API traffic on the designated service port, port 8000
- Launch steps: Start the AMI on a supported GPU instance, allow 5 minutes for the service to fully launch, and begin sending audio via the streaming or batch API
Data Handling and Security
- All audio processing occurs locally on the EC2 instance within your VPC
- No audio data or transcription results are transmitted to external servers
- Supports encryption at rest via AWS EBS encryption
- No telemetry or usage data is sent outside your environment
Highlights
- Single Multilingual Model: Transcribe speech across 40 language-locales using one multilingual model, with optional automatic language detection.
- Punctuation and Capitalization: Automatically produces transcripts with proper punctuation and capitalization for polished, readable output.
- Cost Efficient and Performant: Optimized to maximize GPU performance, allowing more concurrent transcription sessions per instance and helping organizations reduce the cost of large-scale speech recognition deployments.
Details
Sold by
Categories
Delivery method
Delivery option
64-bit (x86) Amazon Machine Image (AMI)
Latest version
Operating system
Ubuntu 26.04
Deployed on AWS
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
Free trial
Try this product free for 5 days according to the free trial terms set by the vendor. Usage-based pricing is in effect for usage beyond the free trial terms. Your free trial gets automatically converted to a paid subscription when the trial ends, but may be canceled any time before that.
Pricing is based on actual usage, with charges varying according to how much you consume. Subscriptions have no end date and may be canceled any time.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimension | Cost/hour |
|---|---|
g4dn.xlarge Recommended | $0.09 |
g4dn.2xlarge | $0.09 |
g4dn.4xlarge | $0.09 |
g4dn.8xlarge | $0.09 |
Dimensions summary
You pay an hourly software fee for this speech-to-text model, which runs on a GPU instance in your own AWS account. Pricing is billed by the hour, not by usage volume or tokens. The four dimensions map to four GPU instance sizes: g4dn.xlarge, g4dn.2xlarge, g4dn.4xlarge, and g4dn.8xlarge. Each step reflects more compute capacity on the underlying instance. You choose the instance size that fits your workload. The model stays in your account, and the hourly fee covers unlimited use of that single instance.
Top-of-mind questions for buyers
Each dimension matches one GPU instance size, from g4dn.xlarge up to g4dn.8xlarge. Larger sizes carry more compute and a higher hourly fee. You pick one size and pay its hourly software fee. The sizes do not stack or combine on a single bill.
The hourly software fee meters running time on the instance you launch. A fully stopped instance does not accrue the software fee. Stopped instances may still incur underlying AWS storage charges for the attached volume, which are separate from this listing's fee.
You pay an hourly software fee for the running instance, not per audio minute, token, or request. Transcription volume on that instance is unlimited while it runs. This favors steady workloads where the instance stays active. Continuous use spreads the hourly fee across more work.
www.salientengineering.com
Helpful?
Vendor refund policy
Refunds may be considered on a case-by-case basis.
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
64-bit (x86) Amazon Machine Image (AMI)
Amazon Machine Image (AMI)
An AMI is a virtual image that provides the information required to launch an instance. Amazon EC2 (Elastic Compute Cloud) instances are virtual servers on which you can run your applications and workloads, offering varying combinations of CPU, memory, storage, and networking resources. You can launch as many instances from as many different AMIs as you need.
Version release notes
Production-ready release.
Additional details
Usage instructions
Once deployed, ensure port 8000 is allowed in the EC2 instance's security group. Then, replace the public DNS endpoint and sample audio file in the following curl command and run it to test the service:
curl http://ec2-123-123-123-123.compute-1.amazonaws.com:8000/v1/audio/transcriptions -F file=@TestAudio.mp3 -F model=nemotron-3.5-asr-streaming-0.6b
The exposed API endpoint adheres to the OpenAI spec and can be used as such.
Support
Vendor support
Getting Help
For deployment support, configuration assistance, performance tuning, and troubleshooting, contact the support team at: support@salientengineering.com
What Support Covers
- Initial deployment and AMI launch guidance
- Performance optimization and GPU utilization tuning
- Troubleshooting transcription errors or service health issues
Getting Started
After launching the AMI on a supported GPU instance, wait 5 minutes then verify the service is running by checking the health endpoint. The ASR API accepts audio input via streaming or batch modes.
Requesting Refunds
For billing questions or refund requests related to AWS Marketplace charges, contact support@salientengineering.com with your AWS account ID and subscription details.
Please include your AWS instance ID, region, and a description of the issue when contacting support to help expedite resolution.
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
Deepgram is the enterprise Voice AI platform for building and scaling real time voice applications on AWS. This product listing contains multiple versions of the nova-3 model which can each transcribe a set of languages. See version details for more information. Deepgram charges are billed per request as described by https://deepgram.com/pricing
Our APIs for Nova Speech to Text (STT) are natively available in the new SageMaker Bi-Directional Streaming API. Additional native touchpoints with Amazon Bedrock, Lex, and Amazon Connect make it simple to compose full voice experiences with the cloud services your teams already trust.
Deepgram is the enterprise Voice AI platform for building and scaling real time voice applications on AWS. This product listing contains multiple versions of the nova-3 model which can each transcribe a set of languages. See version details for more information.
You will be billed $0.0092/min as described by https://deepgram.com/pricing. Private pricing available upon request.
Our APIs for Nova Speech to Text (STT) are natively available in the new SageMaker Bi-Directional Streaming API. Additional native touchpoints with Amazon Bedrock, Lex, and Amazon Connect make it simple to compose full voice experiences with the cloud services your teams already trust.
Deepgram is the enterprise Voice AI platform for building and scaling real time voice applications on AWS. This product listing contains multiple versions of the flux model which can each transcribe a set of languages. See version details for more information.
You will be billed $0.0077/min as described by https://deepgram.com/pricing. Private pricing available upon request.
Our APIs for Nova Speech to Text (STT) are natively available in the new SageMaker Bi-Directional Streaming API. Additional native touchpoints with Amazon Bedrock, Lex, and Amazon Connect make it simple to compose full voice experiences with the cloud services your teams already trust.
Deepgram, Language AI models to power your apps.
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.