Amazon Polly vs Cartesia: pricing, features and which to choose

Amazon Polly launched at AWS re:Invent in November 2016 and is one of the oldest cloud text-to-speech services still in active development. It sits alongside Amazon Transcribe and Amazon Lex in the AWS AI services group, and it powers voices in Alexa features, Amazon Connect contact centers and thousands of third-party apps. Polly has no studio or editor of its own. You call it from the AWS console, the command line, an SDK or another AWS service such as Lambda, and it returns an MP3, OGG or PCM audio stream. Because it is part of AWS, it inherits the platform's compliance coverage, regional data residency, IAM permissions and consolidated billing. That makes it a default choice for teams whose infrastructure already runs on AWS and who want speech without adding another vendor.

A request sends plain text or SSML, a voice ID, an engine and an output format. SSML tags control pauses, emphasis, pronunciation, speaking rate and pitch, and Polly adds its own tags for breathing sounds and a newscaster speaking style on some neural voices. Lexicons let you define how brand names or acronyms should be pronounced across every request. Speech Marks return timing metadata for each word, sentence or viseme, which apps use to highlight text in sync with audio or drive an avatar's lip movements. For long content, asynchronous synthesis tasks write the output to an S3 bucket instead of streaming it. Audio can be cached and replayed at no extra charge, which keeps costs down for apps that repeat the same prompts.

Polly's weakness is voice quality at the top end. Its generative voices are good, but independent rankings such as the Artificial Analysis TTS leaderboard place ElevenLabs, Cartesia and newer models ahead on naturalness, and Polly has no instant voice cloning. Google Cloud Text-to-Speech offers more languages and Chirp 3 HD voices at similar prices, while Azure AI Speech has the largest neural voice catalogue and a custom neural voice program. For a developer already on AWS, Polly's strengths are the 5 million free standard characters, IAM-based security, S3 output and predictable per-character billing with no seats or plans. Choose Polly for IVR prompts, accessibility features, notifications and high-volume narration inside AWS. Choose ElevenLabs or Cartesia when the voice itself is the product.

Cartesia is a San Francisco startup founded in 2023 by Karan Goel, Albert Gu, Arjun Desai, Brandon Yang and Chris Ré, researchers from Stanford's AI lab. Gu co-created the S4 and Mamba state space model architectures, which process long sequences with less memory than transformers, and that research is the basis of the company's products. Cartesia launched its Sonic text-to-speech model in 2024 and raised a $64 million Series A led by Kleiner Perkins in March 2025, bringing total funding to $91 million, according to Fortune. Its homepage names ServiceNow, Retell, Blue Machines and Quora among customers. The company's pitch is speed: models that start speaking quickly enough for live phone conversation, sold through one API that covers speech generation, transcription and complete voice agents.

Developers reach Cartesia through a REST and WebSocket API or the browser playground. Sonic turns text into streaming audio, with controls for speed, emotion and pronunciation, and supports more than 40 languages. Instant voice cloning creates a voice from a few seconds of audio on paid plans, and professional cloning trains a closer copy from longer recordings on the Startup plan and above. A localisation feature gives a cloned voice a new accent. Ink transcribes live audio for agents that need to hear callers. Managed Agents, the newest product, hosts complete phone agents with telephony, so a team can define an agent's prompt and tools without wiring a separate orchestration layer. Cartesia also offers on-premises and on-device deployment for customers with data residency rules.

Cartesia's voice library and creator tools are thinner than ElevenLabs, which also offers dubbing, sound effects and music, so it is a weaker fit for video producers and audiobook narrators. Managed Agents are newer than Vapi or Retell and have fewer integrations. Credit-based billing means long-form narration gets expensive faster on the low tiers. ElevenLabs Flash competes directly on latency with a larger voice catalog. Deepgram's Aura voices are cheaper per minute but less expressive, and Hume, once a rival on expressive speech, is closing its TTS API in November 2026. Pick Cartesia when you are building a real-time voice agent or interactive app and the delay before the first syllable is the metric you care about, or when you need on-device or on-premises speech models.