Skip to main content
Eclatira is a platform for voice agents. You create an agent once. It can then talk to people in two places:
  • In the browser. The user speaks into their microphone. They can also share their camera or screen. The agent answers out loud in real time.
  • On the phone. The agent places outbound calls. You can call one number or a whole list.
One REST API controls both. You authenticate with an API key that starts with ek_.

Quickstart

Get an API key. Then talk to an agent in the browser or place your first call.

API reference

Every endpoint, with request and response examples you can run.

​
How it works

1

Create an agent

An agent has instructions, a voice and a first message. Create it in the dashboard or with POST /api/v1/agents.
2

Give it tools and knowledge

Attach tools, MCP servers and knowledge base files. The agent can use them during a conversation.
3

Put it in front of people

Start a web session from your site, or place a phone call through the API.

​
Choose how to build

Embed widget

Paste two script tags. You get voice, camera and screen share. You write no code.

Client library

Use @eclatira/realtime to put the agent inside your own UI. It handles all the audio and video work.

Phone calls

Place one call or a batch of calls with a single REST request.
Not sure which to pick? Start with the embed widget. Move to the client library when you need the conversation data inside your app.

​
Explore the API

Agents

Create and configure agents.

Calls

Place calls and read results.

Batches

Call a list of numbers.

Realtime

Start a browser session.

Tools

Let agents take actions.

MCP servers

Connect agents to MCP tools.

Knowledge base

Upload files for agents to use.

Webhooks

Get notified when a call ends.

API keys

Create and revoke keys.