Hexabot – Self-Hostable AI Chatbot Platform

TL;DR

TL;DR: Hexabot is an open-source AI chatbot platform you can self-host — featuring a visual workflow builder, multi-channel support, MCP tool integration, and TypeScript-native extensibility.

Source and Accuracy Notes

• Official site: hexabot.ai
• Documentation: docs.hexabot.ai
• GitHub: github.com/Hexastack/Hexabot
• Discord community: discord.gg/rNb9t2MFkG

What Is Hexabot?

Hexabot v3 is an open-source AI chatbot and workflow automation platform built with TypeScript/Node.js. It positions itself as a self-hosted alternative to hosted services like Intercom or Drift, targeting teams that need data residency, custom AI tool integrations, or white-label deployment.

Key capabilities:

• Visual workflow builder — design conversation flows with a YAML-based or CLI-driven approach
• Multi-channel — supports web chat, WhatsApp, and Discord out of the box
• MCP tool integration — connect to external tools via the Model Context Protocol
• Memory and RAG — built-in context management for multi-turn conversations
• Self-hosted — Docker-based deployment, no vendor lock-in
• Extensions — plugin system for custom channels, AI providers, and storage backends

Setup Workflow

Step 1: Install the CLI

npm install -g @hexabot-ai/cli

Or run without a global install:

npx @hexabot-ai/cli --help

Step 2: Create a new project

hexabot create my-hexabot
cd my-hexabot

The CLI auto-detects your package manager (npm, pnpm, yarn, bun). You can force one:

hexabot create my-hexabot --pm npm

Step 3: Start the development server

hexabot dev

This starts the admin UI and API locally. Default endpoints:

• Admin UI: http://localhost:3000
• API: http://localhost:3000/api

Step 4: Configure a channel

In the admin UI, navigate to Channels and enable a channel (web, WhatsApp, or Discord). Each channel has its own webhook configuration.

For Discord, you need a bot token from the Discord Developer Portal.

Step 5: Connect an AI provider

Navigate to Settings → AI Providers. Hexabot supports multiple LLM backends. You’ll need an API key for your chosen provider.

Step 6: Docker deployment (production)

docker compose up -d

The Docker setup includes the API, admin UI, and a PostgreSQL database.

Deeper Analysis

Architecture

Hexabot is a monorepo built around a core runtime and a plugin system. The workflow engine processes conversation states and triggers actions defined in YAML files. MCP integration allows external tools to be invoked mid-conversation.

Strengths

• True self-hosting — full data control, no SaaS dependency
• Visual + code-first — YAML workflows are readable and version-controllable
• MCP native — built-in support for the Model Context Protocol standard
• Active development — v3 is the current major version with regular releases

Limitations

• CLI-first setup — requires terminal access; not a pure click-and-run SaaS
• Documentation gaps — some advanced features (RAG tuning, custom extensions) have thin coverage
• Node.js only — no native Python support for custom plugins

Practical Evaluation Checklist

• [ ] Successfully install CLI and create a project
• [ ] Start dev server and access admin UI
• [ ] Create a simple conversation flow
• [ ] Connect a channel (web or Discord)
• [ ] Configure an AI provider and test a conversation
• [ ] Deploy via Docker and verify data stays local

Security Notes

• Self-hosting keeps conversation data on your infrastructure — no third-party data processing
• Use environment variables for API keys and database credentials
• Enable TLS/SSL in production Docker deployments
• Review Docker Compose file for exposed ports before exposing to the internet

FAQ

Q: Can I use Hexabot without coding knowledge? A: Partially. The visual workflow builder and admin UI handle most configuration tasks. Custom plugins or advanced integrations require TypeScript/JavaScript knowledge.

Q: How does Hexabot compare to Botpress or Voiceflow? A: Hexabot is open-source and self-hostable, while Botpress and Voiceflow are primarily SaaS with some self-hosted options. Hexabot’s MCP integration is a differentiator for AI-native workflows.

Q: Does it support voice channels? A: Not natively in v3. Text channels (web, WhatsApp, Discord) are the primary supported channels. Voice could be added via a custom extension.

Q: What AI models does it support? A: Hexabot v3 supports multiple LLM backends via a configurable AI provider system. Check the documentation for the current list of supported providers.

Conclusion

Hexabot is a solid choice for teams that want an open-source, self-hostable chatbot platform with modern AI features — particularly MCP tool integration and multi-channel support. The CLI-driven setup and TypeScript foundation will feel natural to developer teams, while the admin UI keeps non-technical operators productive. If you need full data control and custom AI workflows without SaaS lock-in, it is worth evaluating.

Next steps:

• Try the quick-start with npx @hexabot-ai/cli create my-hexabot
• Read the full docs at docs.hexabot.ai
• Join the Discord community for support and extension sharing