Amazon Polly
Amazon Polly
Cloud text-to-speech API with pay-per-character pricing
Overview
Amazon Polly is a cloud text-to-speech service from Amazon Web Services. Developers send text to the service and receive spoken audio, which can be streamed or saved as a file. It is built for applications such as read-aloud features, e-learning, voice assistants, accessibility tools and automated announcements, and it is used through the AWS console, the command line, SDKs and an API rather than through a consumer app.
Polly is sold by voice engine, and the engine decides the price. The pricing page lists Standard, Neural, Generative and Long-form engines, with Standard the cheapest and Long-form the most expensive. Billing is based on the number of characters sent for synthesis, and the page states the price per one million characters for each engine. A free tier covers the first 12 months of a new AWS account, with different monthly character allowances per engine, so small projects can test the service before paying.
Because it is an infrastructure service, there is no studio editor for non-technical users. Teams that need a simple web tool for pasting text and downloading audio may find it more work to set up than a consumer product. Usage is billed through an AWS account, and charges depend on volume, so budgets should be set up in AWS before heavy use.
At a glance
It is not a good fit for people who want a point-and-click studio, because there is no consumer editor. Costs scale with the number of characters, so large volumes need monitoring, and the free tier only lasts 12 months.
Key Features
• Billing per character with a stated price per 1 million characters
• Speech returned as streamed audio or as a file
• Access through the AWS console, command line, SDKs and API
• A 12-month free tier for new AWS accounts
• Usable for read-aloud, e-learning, accessibility and announcements
Visual Walkthrough
Best For
• Teams that already use AWS
• Organizations that need to convert large amounts of text at a known per-character rate
• Creators who need voice cloning of their own voice
• Anyone who needs a fixed monthly price, since billing is by usage
Use cases
• Voice prompts in apps and phone systems
• E-learning narration
• Accessibility features
• Automated announcements
Pricing
Standard: $4 per 1 million characters
Neural: $16 per 1 million characters
Generative: $30 per 1 million characters
Long-form: $100 per 1 million characters
Free tier: included for the first 12 months of a new AWS account, with allowances that differ by engine
Prices and terms may change. Check the official site for current information.
Pros and cons
• Several voice engines to match budget and quality
• Works through console, SDKs and API
• Backed by a large cloud provider
• Free tier for new accounts
• No consumer studio or editor
• Usage-based billing can grow with volume
• The free tier is limited to the first 12 months
• Higher-quality engines cost much more than Standard
Integrations
FAQ
A: By the number of characters, with a different price per 1 million characters for Standard, Neural, Generative and Long-form engines.
Q: Is there a free tier?
A: Yes. The pricing page lists a free tier for the first 12 months of a new AWS account.
Q: Is there a web app?
A: Polly is used mainly through the AWS console, command line, SDKs and API.
Q: Who is it for?
A: Developers and teams that need to turn text into speech in their own products.
Alternatives
NaturalReader: text-to-speech reader apps for personal use.
WellSaid: studio-style voices for business narration.
Cartesia: developer-focused voice API with instant cloning.












