Kelu vs Botpress
Last reviewed:
Botpress is a platform for building AI agents: in Botpress Studio, a visual builder, you assemble flows, knowledge and integrations, and Botpress now also sells its own support helpdesk, Botpress Desk. Kelu is a packaged AI knowledge assistant: you connect docs, and the widget, bots, MCP, retrieval and analytics are ready. This comparison is mostly about a product vs a builder.
Choose Botpress if you're building an agent with branching flows and custom actions, on channels like WhatsApp, Telegram or Instagram, and you have time to design it. Choose Kelu if the job is answering questions from your docs and support content, with citations and an abstention gate already built in, on a widget, Slack, Teams, MCP or your helpdesk.
Feature comparison
Botpress details come from Botpress's own website, pricing page and documentation, listed under the table.
Where Kelu focuses differently
A product, not scaffolding
Botpress is powerful when you want to build a custom agent with branching logic, custom actions, and channels like WhatsApp or Instagram. If your use case is 'answer questions from our docs', Kelu is that assistant already: you skip the flow-building and the retrieval tuning.
Retrieval tuned for docs
Kelu's ingest pipeline knows what a docs page looks like: it strips navigation and boilerplate, extracts code blocks, chunks by heading, and finds headings in PDFs. Retrieval combines vector and full-text search with reranking, and refuses to answer when your sources don't cover the question. In a builder, that accuracy work is yours.
Nothing to design before the first answer
Add a source URL, wait for the crawl to finish, and embed the widget or install the Slack or Teams app. There are no flows to design and no prompts to wire up before the first cited answer.
MCP and form deflector included
The MCP server (for Cursor, Claude Desktop, ChatGPT and Windsurf) and the support form deflector are part of the product, on every plan, not something you assemble.
Got questions?
Frequently asked questions
When would I still choose Botpress?
If you're building an agent that lives on WhatsApp, Instagram or Messenger with complex branching logic and custom actions against your own systems, where 'answer from docs' is only part of the job, a builder gives you that flexibility. Botpress Desk is also an option if you want a helpdesk from the same vendor. For docs and support answers on a widget, Slack or MCP, that flexibility is overhead.
Does Kelu let me customise the assistant's behaviour?
Yes: a persona prompt, response tone and custom instructions per knowledge base, reusable customizations you can assign per surface, version labels and (on Enterprise) source groups to scope answers, and theme settings and per-component styles on the widget. What you can't do, on purpose, is rebuild the retrieval or answer pipeline; that's the packaged part.
How long does it take to be live?
Sign up, connect a docs source, wait for the crawl and embed the widget. A first cited answer takes minutes once the source has synced. How long a Botpress agent takes depends on the flows and knowledge you build.
How does pricing compare?
Kelu has flat monthly pricing with included quotas (Free / $99 Pro / Enterprise). Botpress prices per conversation with AI usage included: a free plan with 100 conversations a month, then paid plans such as Plus at $150/month billed annually for 250 conversations, with extra packs on top.
Do I lose flexibility by using a packaged product?
For flow-based bots, yes. For docs and support Q&A, no: the REST API, MCP endpoint and WebSocket streaming give you enough surface to build custom experiences on top of the packaged assistant.
See for yourself
Connect your docs and get your first AI-powered answer in under 5 minutes. No credit card required.
Free to start · No credit card required · Talk to Sales →