The short answer
Cartesia is billed per character (credits). It has a free tier: 20k credits (~15-20 min audio). The paid plans we list start at $5/mo (Pro).
Plans and prices
| Plan | Price | What it covers |
|---|---|---|
| Free | $0/mo | ~20,000 model credits (~27 min TTS), 1 agent slot, basic TTS + STT, no time limit. |
| Pro | $5/mo | ~133 min TTS, 3 agent slots, adds commercial-use license and instant voice cloning. |
| Startup | $49/mo | ~1,667 min TTS, 5 agent slots, adds professional voice cloning and org support. |
| Scale | $299/mo | ~10,667 min TTS, 10 agent slots, priority support and high concurrency limits. |
| Enterprise | Custom | Volume pricing, custom terms, DPAs/BAAs, SSO, compliance and security review; on-prem/on-device options. |
How Cartesia bills you
The pricing model is per character (credits). Low-latency (sub-100ms) real-time TTS challenger built for voice agents, with clean Fern-generated SDKs and a public status page, but a young track record. Best suited to: real-time tts for voice agents.
The pricing page we read
Source: https://www.cartesia.ai/pricing. Prices change, so confirm there before you commit.
Compare Cartesia with alternatives
Other text-to-speech apis: ElevenLabs, Google Cloud Text-to-Speech, Amazon Polly, OpenAI TTS, Azure AI Speech.
Looking at quality rather than price? See the Cartesia score on APIbenchmarks. Full entry: Cartesia pricing.