# Cartesia > Low-latency speech models and voice agents for developers Cartesia is aimed at developers building real-time voice products rather than at people making voiceovers. It sells the Sonic text-to-speech and Ink speech-to-text models through an API, with voice cloning, and Managed Agents for voice agents that answer and place phone calls. Plans are a monthly pool of credits plus prepaid dollars for agent minutes. - **Category:** AI Voices & Text to Speech - **Made by:** Cartesia - **Website:** https://cartesia.ai - **Runs on:** Web - **Facts last checked:** Oct 18, 2026 ## Pricing Cartesia has a free plan. Paid plans start at $5 per month for 100,000 credits, billed monthly. - **Pricing model:** Free plan - **Free plan:** 20,000 credits a month (about 27 minutes of text to speech or 1 hour 51 minutes of speech to text) and $1 of voice agent usage. No commercial use license. - **Free trial:** None listed. | Plan | Price | What you get | | --- | --- | --- | | Free | $0 | 20,000 credits a month, about 27 minutes of text to speech, plus $1 of prepaid agent usage. Not licensed for commercial use. | | Pro | $5 per month for 100,000 credits, billed monthly | About 133 minutes of text to speech a month, $5 of agent usage, a commercial use license and instant voice cloning. | | Startup | $49 per month for 1.25M credits, billed monthly | About 1,667 minutes of text to speech a month, $49 of agent usage and 2 Professional Voice Clone slots. | | Scale | $299 per month for 8M credits, billed monthly | About 10,667 minutes of text to speech a month, $299 of agent usage, 4 Professional Voice Clone slots, priority support and higher concurrency. | | Enterprise | Custom | Volume pricing, custom concurrency, DPAs and BAAs, SSO, a shared Slack channel and on-premise or private cloud deployment. | Read from https://cartesia.ai/pricing on Oct 18, 2026. Prices change, so confirm with Cartesia before buying. ## Ways to pay less - **Startups Grant: 12 months of credits** (Startups): Startups at any stage can apply for 12 months of Scale-level usage: 8 million credits a month plus $299 a month for Managed Agents. Students are considered case by case. Source: https://cartesia.ai/startups ## Key features - **Sonic text to speech:** A streaming text-to-speech model built for live conversation, with dozens of languages and regional accents. One minute of audio uses about 750 to 800 credits. - **Ink speech to text:** A streaming transcription model with its own concurrency pool, so heavy transcription does not slow text to speech. - **Voice cloning:** Instant clones from 10 seconds of audio start on Pro; Professional clones start on Startup. Cartesia says to clone only your own voice or one you have explicit permission to use. - **Managed Agents:** Build voice agents that answer and make phone calls, with Cartesia-provided phone numbers, from 1 on Free to 10 on Scale. - **Deployment options:** Call the models through regional API endpoints with Python and JavaScript SDKs, or arrange on-premise, private cloud or on-device deployment with sales. ## Good for - Developers building real-time voice agents and phone bots - Apps that need fast streaming speech and transcription through one API - Startups that can use the 12-month credit grant ## Worth knowing - Cartesia is a developer platform: you use it through the API or a web playground, and there is no voiceover editor or mobile app. - The Free plan has no commercial use license; it starts on Pro, which also adds instant cloning. Professional cloning needs Startup or above. - Paid plans can go past their credits at $65 per million on Pro, $45 on Startup and $38 on Scale; with overages off, requests are blocked. - Voice agent calls cost $0.06 a minute from the plan's prepaid dollars, plus $0.014 a minute on a Cartesia phone number. ## ToolGPA grades **GPA: 2.4 out of 4**, graded Oct 18, 2026. Cartesia is for developers building voice agents and phone bots, not for people making voiceovers by hand, so we graded it against its alternatives. Its speech models stream fast, cloning starts on the $5 plan, and that plan covers about 133 minutes of speech. But there's no voiceover editor and no app: you work through the API or a web playground. - **Value for money: B.** Pro is $5 a month for 100,000 credits, about 133 minutes of speech, with a commercial license, instant cloning and $5 of agent calls. An hour a month fits on the cheapest paid plan, which undercuts its alternatives for the same amount. - **Free plan and trial: C.** 20,000 credits a month, about 27 minutes of speech, plus $1 of agent usage. Plenty for testing, but there's no commercial use license, so nothing built on it can go live. - **Features: C.** Streaming text to speech in dozens of languages, streaming transcription, instant and professional cloning, and voice agents with phone numbers. It does the developer job fully, but that's a narrower job: there's no voiceover editor, as the listing says. - **Ease of use: C.** No store rating on record. It's built for developers: you call it from code with the Python or JavaScript SDK, or try voices in a web playground. Fine if you code, slow going if you don't. - **Platforms and integrations: B.** Web and an API with regional endpoints, plus LiveKit, Pipecat, Rasa and Genesys for voice agents, and on-premise setups through sales. No phone or desktop app. These are ToolGPA's own grades. Graded from this listing, whose facts were read from Cartesia's own pages on 18 October 2026. Not based on using the product, and shorter than our full reviews. ## Where it stands among AI voice generators - Cartesia's GPA of 2.4 is higher than 4 of the 9 other AI voice generators we grade, and level with 2. - It stands best on value for money (B, above 9 of 9) and worst on free plan and trial (C, above 0). - Its paid plans start at $5 a month, against a middle price of $14 for the category: 0 of the 9 others that charge start lower. - On features, ease of use and platforms it scores 2.3 out of 4. Nothing scores higher for less, though at least one tool matches that score at a lower price. The category's chart of price against score: https://toolgpa.com/charts/ai-voices ## Keep it, or make your own? **Keep it.** To replace Cartesia you would be making a voice model, which you can rent by the minute but not make. That is not a weekend's work, with an AI app builder or without one. It costs $60 a year. - You could not: A voice model, which you can rent by the minute but not make. - You could not: A voice model trained on your recordings. - You could not: Speech turned into text you can trust. - You could not: It plugs into other products, LiveKit, Pipecat and Rasa among them, and has an API. A judgement by ToolGPA from this listing, by rules that are the same for every tool. Nothing was built to test it. ## Works with LiveKit, Pipecat, Rasa, Genesys **Tags:** Text to speech, Voice cloning, Transcription, Public API ## Alternatives - [ElevenLabs](https://toolgpa.com/tools/elevenlabs): AI voice platform for speech, voice cloning, dubbing and voice agents. [Compare](https://toolgpa.com/compare/cartesia-vs-elevenlabs) - [Murf](https://toolgpa.com/tools/murf): AI voiceover studio with dubbing, translation and a speech API. [Compare](https://toolgpa.com/compare/cartesia-vs-murf) - [Typecast](https://toolgpa.com/tools/typecast): AI voice generator with emotion controls, voice cloning and video. [Compare](https://toolgpa.com/compare/cartesia-vs-typecast) - [Respeecher](https://toolgpa.com/tools/respeecher): Voice cloning and speech-to-speech for film, TV, games and developers. [Compare](https://toolgpa.com/compare/cartesia-vs-respeecher) ## Sources The prices, plans, platforms and features above were read from these pages on Oct 18, 2026: - https://cartesia.ai - https://cartesia.ai/pricing - https://cartesia.ai/startups - https://cartesia.ai/product/voice-cloning - https://cartesia.ai/languages - https://docs.cartesia.ai/llms.txt Page: https://toolgpa.com/tools/cartesia Listed by ToolGPA (toolgpa.com).