Build your first verified voice agent

Learn how to make your first request through Cipher's Proxy Platform and connect your voice provider. The proxy is fully OpenAI-compatible and streams responses, so you can drop it into existing integrations.

Getting started

Follow these steps to integrate Cipher with your voice agent platform

Create an API key

.env

# Add to your environment
CIPHER_API_KEY=sk_live_your_key_here

Configure your downstream model

Cipher forwards verified traffic to your preferred LLM provider and model. You can configure this in the dashboard, or via API:

curl -X POST https://api.getcipher.ai/proxy/model \
  -H "Authorization: Bearer $CIPHER_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{"model": "gpt-4o-mini"}'

Popular models include gpt-4o-mini, gpt-4o, and gemini-2.5-flash. Choose based on your latency and quality requirements.

Make a test request

The proxy speaks OpenAI's Chat Completions API. Point your request at https://api.getcipher.ai/v1/chat/completions with your Cipher API key. If using phone verification, include the X-Office-Number header.

curl -N https://api.getcipher.ai/v1/chat/completions \
  -H "Authorization: Bearer $CIPHER_API_KEY" \
  -H "Content-Type: application/json" \
  -H "X-Office-Number: +14155551212" \
  -d '{
    "model": "gpt-4o-mini",
    "stream": true,
    "messages": [
      {"role": "system", "content": "You are a helpful agent."},
      {"role": "user", "content": "Say hello!"}
    ]
  }'

Connect your voice provider

Replace your provider's LLM URL with Cipher and keep streaming enabled. Use your Cipher API key as the bearer token.

Vapi Integration

Create or edit an Assistant in Vapi
Set LLM Provider to Custom/OpenAI-compatible
Base URL: https://api.getcipher.ai/v1
Endpoint/Path: /chat/completions
Auth: Authorization: Bearer $CIPHER_API_KEY
(Optional) Header: X-Office-Number set to the caller's number

Vapi LLM configuration

View detailed Vapi setup guide →

ElevenLabs Agents Integration

Open your Agent or Workspace settings in ElevenLabs
Set the LLM provider to Custom/OpenAI-compatible
Custom LLM URL: https://api.getcipher.ai/v1/chat/completions
Auth: Authorization: Bearer $CIPHER_API_KEY
(Optional) Header: X-Office-Number to enable phone verification

ElevenLabs LLM configuration

View detailed ElevenLabs setup guide →

Troubleshooting

Common issues and how to resolve them

Missing stream? Ensure stream=true is passed in your request and your provider supports Server-Sent Events (SSE).

401 errors? Double-check the Authorization header format. You can use either Authorization: Bearer $CIPHER_API_KEY or the custom header x-cipher-api-key.

Phone verification not triggering? Include the X-Office-Number header with E.164 formatted numbers (e.g., +14155551212).

Next steps

Vapi Setup →

Detailed guide for integrating Cipher with Vapi assistants

ElevenLabs Setup →

Step-by-step instructions for ElevenLabs agents integration

Platform Overview →

Learn more about Cipher's proxy architecture and features

MCP Platform →

Explore Model Context Protocol integration for advanced use cases

curl -N https://api.getcipher.ai/v1/chat/completions \ -H "Authorization: Bearer $CIPHER_API_KEY" \ -H "Content-Type: application/json" \ -H "X-Office-Number: +14155551212" \ -d '{ "model": "gpt-4o-mini", "stream": true, "messages": [ {"role": "system", "content": "You are a helpful agent."}, {"role": "user", "content": "Say hello!"} ] }'