WhatsApp

Deploy agents to WhatsApp Business for customer-facing interactions.

The WhatsApp interface turns any Agno agent, team, or workflow into a WhatsApp bot via the WhatsApp Business API, with media handling and interactive messages.

Quick Start

whatsapp_agent.py

from agno.agent import Agent
from agno.models.openai import OpenAIResponses
from agno.os.app import AgentOS
from agno.os.interfaces.whatsapp import Whatsapp

agent = Agent(name="Assistant", model=OpenAIResponses(id="gpt-5.4"))

agent_os = AgentOS(
    agents=[agent],
    interfaces=[Whatsapp(agent=agent)],
)
app = agent_os.get_app()

uv pip install 'agno[whatsapp]'
python -m agno.os.serve whatsapp_agent:app --port 7777

How It Works

Concept	Behavior
Sessions	Each phone number is a session. Send /new to reset.
Media	Images, video, audio, documents downloaded and passed to agent
Security	HMAC-SHA256 signature validation on all webhooks
Encryption	Optional AES-256-GCM phone number encryption

Setup

Requires a Meta Developer account, WhatsApp Business API access, and webhook configuration. See the setup guide for step-by-step instructions.

Environment variables:

export WHATSAPP_ACCESS_TOKEN="..."
export WHATSAPP_PHONE_NUMBER_ID="..."
export WHATSAPP_VERIFY_TOKEN="..."
export WHATSAPP_APP_SECRET="..."  # production

Next Steps

Full Guide

Sessions, media, security, and troubleshooting.

Reference

All parameters, endpoints, and events.

Examples

Agents, teams, and multimodal examples.

WhatsAppTools

Buttons, list menus, locations, and reactions.