Hexabot

What it is

Hexabot is a self-hosted AI workflow platform (the page calls V3 "available now") where you build conversational agents, agentic workflows, and business automation on infrastructure you control. Unlike cloud-first chatbot builders, Hexabot is source-available under the Fair Core License (v1.0) and ALv2 Future License, meaning you can inspect, self-host, and extend the full codebase. The pricing page speaks to freelancers, indie developers, agencies, implementation partners and regulated teams, and the compare page frames it for engineering teams that need to keep production data, workflow logic and memory under their own control. The site says you can deploy on-premise or in your cloud with full source visibility.

What you build and how

Setup starts from the CLI. Install the official Hexabot CLI, create a project from a starter template, move into the workspace, and run the local stack (hexabot dev). You define workflows visually or as portable YAML with typed contracts and schema validation; workflows are version-controllable and merge-friendly. Workflows trigger from chat, forms, schedules, or events; they then call models, tools, APIs, and back-office systems. The platform emphasizes conversational agents, not just chatbots: agents manage conversation memory, context, and multi-turn interactions.

Actions and bindings connect reusable pieces. You wire up conversations to business logic through actions, LLM calls, capability bindings, and integrations. The visual builder and YAML option let teams choose their workflow style: designers work in the UI, engineers define portable config in version control. Hexabot includes managed conversation memory and RAG (retrieval-augmented generation) out of the box. The site lists RAG next to memory, and the Compare page marks Hexabot as having both, so answers can draw on your own content.

The platform connects to messaging platforms, websites, social channels, and custom entry points via deployable channels. Support teams manage live interactions centrally and step in when a workflow needs human judgment. Developers can build custom channels. The site does not name the built-in channels beyond messaging platforms, websites and social channels, so ask which ones your project needs before committing.

Hexabot supports MCP (Model Context Protocol) tool integration, meaning you can bind external tools, APIs, and business systems into agent workflows. The platform does not charge by message volume; instead, plans scale on activations (license keys for production deployments), users, and workflow count. The Community plan lets you export workflows and configuration when you are ready to grow. The pricing page states that "message volume is not a billing meter", so a usage spike does not change the price.

The company compares itself to Botpress (cloud-first), Typebot (conversational forms), Voiceflow (design-first, cloud-only), and n8n (general-purpose backend automation). Hexabot positions itself as the self-hosted, workflow-first alternative: source-available, purpose-built for conversational AI agents, with managed conversation memory, RAG, MCP tools, and multi-channel chat. The compare page also says Hexabot gives engineering teams self-hosting, a free Community edition and full data residency control, which it contrasts with Voiceflow, a cloud-only tool in its words.

The evidence from the pages

Hexabot lists 5 plans, all self-hosted: Community (free, 1 user, 10 free workflows, no license key), Starter ($19/month: 3 users, 25 workflows, 3 license activations), Pro ($59/month: 10 users, 150 workflows, 10 activations, plus priority support, roles and longer operational history), Unlimited ($149/month: 25 users, unlimited workflows within fair use, 25 activations, a dedicated success channel), and Enterprise on a custom quote with a dedicated success manager, SLA-backed uptime and compliance support. The pricing page says projects and integration connections are not capped by plan. The platform is source-available under Fair Core License, Version 1.0, ALv2 Future License. Commercial license-key plans provide higher user, workflow, and activation capacity.

The product is described as: "It combines multi-channel conversational agents with managed memory, RAG, MCP tool integration, and visual or YAML workflows" and "a source-available, self-hostable chatbot and AI workflow platform." The pages position it as built for teams that value "full source visibility," want to "control production data, workflow logic, memory, and extensions," and prefer "keeping workflows, memory, tools, and customer data in infrastructure you control."

In comparison tables, Hexabot is marked Yes for: self-hosted/source-available edition, keeping production data under your control, purpose-built conversational AI agents, managed conversation memory plus RAG, MCP tool integration, visual plus YAML workflows, multi-channel deployment, embed/share, and free self-hosted Community edition. Competitors are marked differently: Botpress as cloud-first with self-host available, Typebot as forms-focused, Voiceflow with no self-hosted edition, and n8n as general-purpose with 400+ integrations. The differentiation reflects Hexabot's primary design philosophy: ownership and control of your production runtime.

The site shows 1k+ GitHub stars on its home page and 940+ on the Compare page, and offers a free Community edition. The site links to documentation, a Discord community, a YouTube channel and an extensions page, and the home page walks through a 4-step CLI install. The pricing table clarifies: "Unlimited means fair-use unlimited, not unbounded abuse."

Who it suits and who it does not

Hexabot is the right choice for engineering teams, agencies, implementation partners and regulated teams that must keep customer conversations, memory, workflow logic, and data residency under their own control. Agencies building automation for clients are the stated audience of the Unlimited plan, which the page calls predictable pricing for agencies, implementation partners and high-usage teams, with capacity measured in users, workflows and activations rather than messages. Developers who want to inspect and extend the codebase, customize channels, or integrate specialized tools benefit from source-available architecture. Organizations with data residency requirements can self-host Hexabot on infrastructure they manage, and the Enterprise plan adds compliance support. Teams wary of a cloud-only vendor may also like that workflows can be modeled as portable YAML under version control.

The honest limitations: First, self-hosting means you own the operational burden: deployment, updates, monitoring, and infrastructure costs are your responsibility (or your cloud provider's). Second, Hexabot is workflow-first and engineering-oriented; the compare page describes Voiceflow as design-first for non-technical CX teams, which is a different audience. Third, setup requires CLI familiarity and local development; it is not a "click to deploy" service. Finally, community and lower-tier plans cap workflows and users; scale requires a higher plan or an Enterprise agreement. For solo founders or small teams building a single chatbot, the self-hosting operational overhead and setup friction outweigh the data-control benefits.