Build custom voice agents with our developer API. REST endpoints for management, WebSocket for streaming. Complete reference and examples.
Everything you need to build voice agents
Build your first voice agent in minutes
# JavaScript/TypeScript npm install @edesy/voice-agent # Python pip install edesy-voice-agent # Go go get github.com/edesy/voice-agent-go
import { VoiceAgent } from '@edesy/voice-agent';
const client = new VoiceAgent({
apiKey: process.env.EDESY_API_KEY
});
// Create an agent
const agent = await client.agents.create({
name: 'Support Bot',
sttProvider: 'deepgram',
llmProvider: 'gemini',
ttsProvider: 'elevenlabs',
systemPrompt: 'You are a helpful support agent...',
language: 'en-IN'
});
// Make an outbound call
const call = await client.calls.create({
agentId: agent.id,
to: '+919876543210',
from: '+918765432109',
webhookUrl: 'https://your-server.com/webhook'
});
console.log('Call started:', call.sid);Secure your API requests
Bearer Token Authentication
curl -X GET https://api.edesy.in/v1/agents \ -H "Authorization: Bearer YOUR_API_KEY" \ -H "Content-Type: application/json"
Base URLs
api.edesy.in/v1 - Productionsandbox.api.edesy.in/v1 - Testingws.api.edesy.in - WebSocketRate Limits
Core endpoints
/v1/agentsCreate a new voice agent
/v1/agents/:idGet agent details
/v1/callsInitiate an outbound call
/v1/calls/:sidGet call status and details
/v1/webhooksRegister a webhook endpoint
Native libraries for popular languages
npm install @edesy/voice-agentpip install edesy-voice-agentgo get github.com/edesy/voice-agent-goSign up for a free account to get your API key. Free tier includes 100 minutes/month. API keys are in Dashboard > Settings > API.
REST API is for agent management, initiating calls, and querying status. WebSocket is for real-time audio streaming during active calls.
STT: Deepgram, Google, Azure, Whisper. TTS: ElevenLabs, Azure, Google, PlayHT. LLM: OpenAI, Gemini, Claude, custom models.
Native support for Twilio, Exotel, Plivo, and Vonage. SIP trunking available for custom providers.