Sold by: Sigmodata
Deployed on AWS
Free Trial
Deploy a production-ready, self-hosted text-to-speech server on AWS EC2. Convert text to natural-sounding speech via a web UI and REST API using Chatterbox Turbo. Models and voices are baked into the AMI - no downloads at boot. Runs entirely within your AWS account.
Overview
Image
Synthesis web UI
Image
Sign in
Image
Admin console
This product provides a fully self-hosted text-to-speech server packaged as an Amazon Machine Image (AMI) for AWS EC2. The server converts written text into spoken audio using Resemble AI's Chatterbox Turbo (MIT license) and exposes both a web interface and an HTTP REST API. All processing occurs within the customer's AWS environment. Model weights, bundled speakers, and CUDA libraries are baked into the image; the instance does not contact Hugging Face, ECR, or any other external service after launch.
GPU instances are the supported path. The recommended type is g4dn.xlarge (NVIDIA T4). Larger g4dn, g5, and g6 types are also supported. CPU-only types (m6i.2xlarge / m7i.2xlarge and above) run the same model more slowly. Small burst instances such as t3 are not supported and are refused at boot.
The AMI launches with a guided browser setup wizard over HTTPS (port 443). You confirm ownership with the EC2 instance ID, choose an administrator password, and pick a certificate option. No SSH or user-data editing is required for credentials.
After setup completes, Chatterbox Turbo loads onto the GPU from weights baked into the AMI. On a recommended g4dn.xlarge instance, plan for about 3 minutes on first launch from a new EBS volume (cold snapshot read). The web UI and GET /health show loading progress (step and progress_pct) through library import, reading weights from disk, GPU load, and a short warmup synthesis. The progress bar can sit at a low percentage while EBS reads the AMI snapshot for the first time. Reboots on the same volume are usually faster. Nothing is downloaded from external services during startup.
Once ready, English synthesis on a g4dn.xlarge (NVIDIA T4) is faster than realtime for typical sentence-length prompts (about 2.5 s wall clock for a one-line request that produces 5 s of audio). Very short phrases are near realtime. The engine runs one GPU synthesis at a time; parallel requests queue.
Common use cases include IVR and telephony prompts, accessibility applications, content narration, automated announcements, and private voice generation pipelines. Optional zero-shot voice cloning accepts a short reference clip (>5 seconds) on the synthesis API.
Highlights
- Turn text into natural speech for IVR, narration, and voice apps without a third-party TTS API
- Chatterbox Turbo with bundled speakers and zero-shot cloning; air-gapped after launch, no weight downloads
- About 3 min first boot on g4dn.xlarge; English synthesis faster than realtime on T4 after warmup
Details
Sold by
Categories
Delivery method
Delivery option
64-bit (x86) Amazon Machine Image (AMI)
Latest version
Operating system
AmazonLinux 2023
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. Alternatively, you can pay upfront for a contract, which typically covers your anticipated usage for the contract duration. Any usage beyond contract will incur additional usage-based costs.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimension | Cost/hour |
|---|---|
g4dn.xlarge Recommended | $0.10 |
c5.large | $0.10 |
c5.xlarge | $0.10 |
c6i.large | $0.10 |
c6i.xlarge | $0.10 |
c7i.large | $0.10 |
c7i.xlarge | $0.10 |
g4dn.2xlarge | $0.10 |
g4dn.4xlarge | $0.10 |
g5.2xlarge | $0.10 |
Dimensions summary
You pay by the hour for the EC2 instance type you run this AMI on. Pricing is not tiered by features; every instance runs the same text-to-speech server. Your rate scales with the compute you pick. GPU instances (g4dn, g5, g6, g6e, p3) are the supported path for speech synthesis. CPU-only types (c5, c6i, c7i, m5, m6i, m7i, r5, r6i) run the same model more slowly. Small burst types (t3, t3a) appear but are refused at boot. Larger sizes within each family carry higher hourly rates for more processing capacity.
Top-of-mind questions for buyers
You pay for each hour the chosen EC2 instance runs. One hour equals one running instance of that type. The rate reflects the compute size you pick. Stopped instances stop the software hourly charge, though AWS storage fees for the EBS volume may still apply separately.
The listing exposes t3 and t3a sizes, but the server rejects these small burst types at boot. They cannot run the text-to-speech workload. Choose a GPU type (g4dn, g5, g6, g6e, p3) for supported synthesis, or a CPU type (c5, c6i, c7i, m5, m6i, m7i, r5, r6i) for slower processing.
No. Your rate is set by the instance type you launch, not by request volume or size. The engine runs one GPU synthesis at a time; extra requests queue on the same instance. To process more in parallel, you launch additional instances, and each one bills its own hourly rate.
www.sigmodata.com
Helpful?
Vendor refund policy
No refunds offered but you may cancel at any time
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
- Boot loading UX: prefetch progress and faster cold start on g4dn.xlarge (~3 min).
- Measured synthesis latency on NVIDIA T4 documented in listing copy.
- AIFF sidecar downloads in batch UI; security hardening on gateway auth path.
Additional details
Usage instructions
- Launch g4dn.xlarge (recommended) with inbound TCP 443 and 22.
- Open https://<instance-address>/ in a browser. The certificate is self-signed at this point, so your browser will warn you once.
- Enter the instance ID shown in the EC2 console to confirm ownership.
- Complete the setup wizard: administrator password and certificate option.
- Wait for Chatterbox Turbo to load onto the GPU. On g4dn.xlarge, plan for about 3 minutes on first launch from a new EBS volume. The web UI shows a loading screen with progress percentage. GET /health (no auth) returns status loading with step and progress_pct until status is ok. Typical steps: import_libs, prefetch_turbo, load_checkpoint, load_voices, warmup. The bar can sit at a low percentage during import_libs and load_checkpoint while EBS reads the AMI snapshot for the first time. Weights come from the baked AMI - nothing is downloaded from the internet. Reboots on the same volume are usually faster.
- Synthesize at https://<instance-address>/ or POST /synthesis with HTTP Basic auth (username admin, password = the wizard password). On g4dn.xlarge after warmup, a one-line English request is about 2.5 s wall clock for 5 s of audio. Parallel requests queue on one GPU.
Port 80 is not required. Let's Encrypt validation uses port 443. Manage the instance later at https://<instance-address>/admin.
Where your data is stored
- Accounts and API tokens: /etc/sigmodata-tts/identity.sqlite3. Passwords and tokens are stored hashed, never in clear text.
- Service configuration: /etc/sigmodata-tts/
- TLS certificate and private key: /var/lib/sigmodata-tts/certs/
- Synthesized audio is returned in the response and is not retained on the
- instance after the request completes. All of it stays on the instance EBS volume in your own AWS account. Nothing is sent to Sigmodata.
Encryption
- In transit: HTTPS on port 443, self-signed by default or Let's Encrypt from the setup wizard.
- At rest: AWS Marketplace requires the published AMI snapshot to be unencrypted. If you need encryption at rest, enable EBS encryption when you launch (Advanced details -> EBS encryption) or launch from an encrypted copy. The product needs no change either way.
Rotation
- The TLS certificate renews automatically on a systemd timer. Renew it or switch certificate type on demand at /admin.
- Revoke and re-issue account passwords and API tokens at /admin at any time. Rotate any credential that may have been exposed.
Checking that the product is healthy
- In the EC2 console, choose Instances, select this instance, and check the Status checks tab for passed system and instance checks.
- Request https://<instance-address>/_gw/healthz for the gateway. GET https://<instance-address>/health returns ok when synthesis is ready, or loading with step and progress_pct while the engine starts (about 3 minutes on g4dn.xlarge first boot from a new EBS volume).
- On the instance, run: sudo systemctl status sigmodata-tts and sudo journalctl -u sigmodata-tts for logs.
Backup and recovery
- Take EBS snapshots of the instance volume for full-instance recovery.
Service quotas and cost
- GPU instances (g4dn/g5/g6) draw on the "Running On-Demand G and VT instances" vCPU quota, which is 0 on a new AWS account. Request an increase in Service Quotas -> Amazon EC2 before you launch.
- You pay the AWS Marketplace hourly charge for this product plus the AWS resources it uses: EC2 instance hours, the EBS root volume, any S3 storage you configure, and data transfer out.
Resources
Vendor resources
Support
Vendor support
Support contact: Email: support@sigmodata.com Support URL: https://www.sigmodata.com
Support description: Support is provided via email and web-based support channels. Buyers can expect assistance with initial deployment, configuration, and troubleshooting of the AMI and text-to-speech service. Response times and support scope may vary based on the customers subscription or private offer agreement.
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
Deploy a production-ready, self-hosted speech-to-text server on AWS EC2. Convert audio into accurate text via a web UI, a REST API, or batch folder conversion over S3, authenticated with your own accounts or API tokens. Uses a GPU automatically when present, falls back to CPU otherwise. All processing stays in your AWS account.
Deploy a production-ready, self-hosted speech-to-text server on AWS EC2. Convert audio into accurate text via a web UI, a REST API, or batch folder conversion over S3, authenticated with your own accounts or API tokens. Uses a GPU automatically when present, falls back to CPU otherwise. All processing stays in your AWS account.
Deepdub eTTS is a cutting-edge neural text-to-speech model delivering ultra-realistic, human-like voices in 100+ languages and accents. Built for AWS SageMaker JumpStart, it enables developers and enterprises to generate expressive speech with natural prosody, emotion, and clarity, directly within their AWS environment. Easily deployable via SageMaker endpoints, Deepdub eTTS supports both streaming and batch workflows, making it ideal for media localization, conversational AI, eLearning, accessibility, and more. With low-latency inference, fine control over tone and style, and seamless AWS integration, Deepdub eTTS empowers you to create lifelike, engaging audio experiences at scale, without compromising on performance or security.
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.