PASAR API

🇲🇾 Cartesia Ultra-Low-Latency TTS & STT

← directory
CartesiaAItier: commercialauth: API keypricing: freemiumverified 2026-06-15

Ultra-low-latency TTS (Sonic 3.5 model, first audio byte in 90ms) and STT (Ink models). Built for real-time voice agents. Free: 10k credits. Supports voice cloning with accent control.

View docs ↗ More AI APIs

Get started

# Cartesia Ultra-Low-Latency TTS & STT needs an API key.
# No single public base URL ~ start from the docs:
# https://docs.cartesia.ai/

AI prompt

I'm integrating a Malaysia API into my app.

API: Cartesia Ultra-Low-Latency TTS & STT
Provider: Cartesia
Auth: requires an API key (sign up: https://docs.cartesia.ai/)
Pricing: freemium
Important note: Ultra-low-latency TTS (Sonic 3.5 model, first audio byte in 90ms) and STT (Ink models). Built for real-time voice agents. Free: 10k credits. Supports voice cloning with accent control.

Source: Cartesia (https://docs.cartesia.ai/)
Source type: Provider API or developer docs
Endpoint type: Docs-first integration
Integration difficulty: Needs API key
Last checked: 2026-06-15
Production caveat: Keep credentials server-side and confirm current limits, pricing, and webhook/error behaviour in the provider docs.

Help me set up the correct onboarding/authentication path and make the first valid request ~ explain exactly where credentials go and what must happen before production.

Related AI APIs

OpenAI API ~ GPT models, embeddings, images, audioAnthropic Claude API ~ Claude LLM messagesGoogle Gemini API ~ multimodal LLMGroq API ~ fast LLM inferenceMistral AI API ~ open and frontier LLMsCohere API ~ LLMs, embeddings, rerankPerplexity API ~ web-grounded answers and searchOpenRouter API ~ one endpoint for many LLMsHugging Face Inference Providers ~ run open modelsReplicate API ~ run and host ML models