Developer documentation
Build voice agents on Aethex.
Build high-quality voice agents. Speech models, infrastructure, and orchestration on us.
Pick a guide below to get started.
Get Started
Build your first voice agent in 5 minutes.
How It Works
Learn the building blocks: agents, voices, and live calls.
API Reference
Complete reference for the AethexAI REST API.
Account
Manage your billing, plan, and usage analytics.
Developer documentation
Build voice agents that get real work done
Everything here is a live API. Compose them into an agent that answers the phone, talks like a person, reaches into your data and tools mid-call, and leaves behind a transcript, a recording, and usage you can read down to the minute.
Build your agent
Shape how your agent thinks, sounds, and stays in bounds.
Agents→Define the role, prompt, and conversation flow.Models→Choose the language model behind replies.Voices→Pick how your agent sounds.Languages→Speak to people in their language.Knowledge base→Ground answers in your own documents.Tools→Let the agent call your API during the call.Personalization→Pass per-call variables into the prompt.Guardrails→Keep the agent on-topic and safe.
Connect and deploy
Put your agent on the phone, the web, and your stack.
WebRTC calls→Talk to your agent from the browser.Phone numbers→Give your agent a number to answer.SIP trunking→Bring your own carrier over SIP.Calls→Have your agent call people, one or in bulk.Integrations→Connect the tools you already run on.SDKs→Python, TypeScript, and React.MCP server→Drive Aethex from AI tools over MCP.Webhooks→Receive call events in real time.
Monitor and improve
See every call, and what it cost.