Skip to content
← Models
ClosedAmazon Web Services/10 trustVerified

Amazon Polly

AWS's production text-to-speech service, now with a generative engine for near-human expressive voices.

Visit official page →

About

Amazon Polly is AWS's long-running text-to-speech service, and the entry that matters now is its generative engine, which went generally available on May 8, 2024 with the voices Ruth and Matthew (US English) and Amy (British English). The engine builds on Amazon's BASE TTS research and is tuned for context-dependent prosody, pausing and pronunciation, aimed at conversational use cases like virtual assistants and trainers as well as long-form narration. It sits alongside the older standard, neural and long-form engines, with 60-plus voices across 30-plus languages.

Pricing is pay-as-you-go per character: $4 per million for standard voices, $16 for neural, $30 for the generative engine and $100 for long-form. New accounts get 5 million standard and 500,000 neural characters a month free for the first year. It integrates with the rest of AWS through boto3 and the other SDKs, and its Speech Marks feature (no extra charge) makes subtitle and lip-sync work straightforward.

Specs

Context window

3,000 characters per realtime SynthesizeSpeech call; up to 200,000 characters via async StartSpeechSynthesisTask

Pricing

Pay-as-you-go: $4/1M chars standard, $16/1M neural, $30/1M generative engine, $100/1M long-form

Free quota

5M standard + 500K neural characters/month free for the first 12 months

API access

API available — SynthesizeSpeech / StartSpeechSynthesisTask via boto3 and other AWS SDKs

Weights

Closed — API only

Released

Not disclosed

Benchmarks

No published benchmarks — quality claims rest on provider positioning and broad production adoption

Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.

Use cases

text-to-speechvoice-agentsaudiobooksaccessibilityivr-systems

Trust score — 9/10

/10 trustVerified

Scored 9/10 (TIER verified, metrics-v1). Evidence: AWS GA announcement (May 8, 2024) for the generative engine; documented per-engine pricing; free-tier terms on AWS; one of the most widely deployed TTS services in production. Strengths: mature service, four engine tiers, generous free tier, deep AWS integration, Speech Marks included. Docked for: (1) closed service with no published quality benchmarks; (2) generative engine limited to US regions at launch and priced 2x above neural. License: proprietary (AWS service).

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
10.0/10
Maintenance · recency, releases, issue load
9.0/10
Openness · license + docs
2.0/10
Manual review · sanity read + red-flag check
9.0/10
  • · Generative engine GA May 8, 2024 (Ruth, Matthew en-US; Amy en-GB); built on BASE TTS research
  • · 60+ voices, 30+ languages; four engines: standard/neural/long-form/generative
  • · Pricing: $4/$16/$30/$100 per 1M chars; 5M standard + 500K neural chars/mo free for 12 months
  • · AWS logo staged

Sources