Botpress

What is Botpress?

Botpress is an open platform for building conversational agents with visual flows and LLM steps. You design an agent in a visual editor, add nodes that call a language model where you need reasoning or generation, connect the agent to channels and data sources, and publish it to a website or app. The result is a bot whose logic you shape and control rather than one you configure from a fixed template.

Botpress the company makes the product. The platform grew from an earlier open-source project into a broader studio for LLM-powered agents, and it keeps an open, developer-first stance with a community and a library of reusable pieces. The builder sits at the center: a developer wires nodes into branching paths and mixes deterministic logic with model-driven steps on one canvas.

The audience is developers who want control over the bot logic. Botpress targets engineers and technical builders who need more than a scripted FAQ bot and want to decide where a language model acts, how data flows, and which channels the agent serves. These teams value the depth of an open platform over the speed of a locked-down template.

Key features

Botpress centers on a set of features that turn model reasoning into a controlled agent:

• ▸Visual flow builder: map an agent node by node on a canvas, wiring branching logic so developers keep control over how the conversation moves.
• ▸LLM nodes: drop language model steps into a flow where you want reasoning, generation, or classification, so the model works inside a structure you define.
• ▸Integrations and channels: connect the agent to web, messaging channels, and outside tools so it acts on data and reaches users where they are.
• ▸Web chat widget: embed the agent on a website as a chat widget that visitors use without leaving the page.
• ▸Knowledge bases: feed the agent documents and content so it answers from your material rather than generic model output.
• ▸Reusable components: build with shared blocks and community pieces that speed up new agents and keep logic consistent.

The mix of visual flows and LLM nodes matters most. Because a developer decides where the model acts and where fixed logic runs, the agent gets the flexibility of a language model without giving up predictable behavior. Integrations and knowledge bases then ground the agent in your channels and data, so it does more than chat: it answers from your content and connects to your stack.

How well does it work?

Botpress performs well at its core job: giving developers a way to build LLM-powered agents while keeping control of the logic. The visual builder makes flow design clear, and LLM nodes let a team place model reasoning where it helps without handing the whole conversation to a black box. Knowledge bases ground answers in your own content, and the wide integration list means an agent can act across channels and tools rather than sit as an isolated widget.

The limits track its developer-first focus. Botpress rewards technical skill, so a non-technical user faces a learning curve that a pure template tool avoids. Getting the most from flows, LLM nodes, and integrations takes engineering time, which small teams may not have. Usage-based pricing means cost scales with message and AI volume, so a busy agent can grow the bill as it grows traffic.

A practical way to get value fast is to build one focused agent for a single use case, such as a support deflection bot on your help pages, and refine its flow and knowledge base before you expand to more channels.

Botpress pricing

Botpress offers a free tier and usage-based paid plans. The free tier lets you build and test an agent with capped messages and AI usage, which suits prototypes and small projects. Paid usage lifts those caps and cost tracks the messages and model spend your agent consumes. Larger organizations move to a custom enterprise plan scoped to volume, security, and support needs.

Here is how the tiers compare so you can match a plan to your stage:

One thing to watch is the usage-based model. Cost tracks message volume and AI consumption, so an agent on high traffic or one that leans on frequent LLM calls can push spend up faster than a flat seat fee would. Model your expected message load and AI usage before you scale a plan.

Who should use Botpress?

Botpress suits developers and technical teams that want to own the logic behind an LLM-powered agent. If you need more than a scripted FAQ bot and want to decide where a language model acts and how data moves, the platform gives you that control across web and app channels.

It fits these cases well:

• ▸Development teams building custom conversational agents that mix fixed logic with LLM reasoning.
• ▸Product teams that want a support or assistant bot grounded in their own knowledge base.
• ▸Businesses that need an agent connected to outside tools and channels rather than a standalone widget.
• ▸Technical builders who want an open platform with reusable components and a community to draw from.
• ▸Teams that value control over the bot's behavior more than the speed of a locked template.

It is a weaker fit for non-technical marketers who want a funnel live in an afternoon, or for teams that prefer a fixed, low-effort template over an open platform they configure themselves.

Alternatives and how it compares

Botpress competes with other bot builders and agent platforms, and the right choice depends on whether your priority is developer control, marketing funnels, or a managed support suite.

• ▸Landbot: a no-code builder for web and WhatsApp lead funnels, faster for marketers but lighter on developer control.
• ▸Voiceflow: an agent design tool with a visual canvas and LLM steps, aimed at collaborative teams building conversational products.
• ▸Intercom Fin: a managed support agent inside a customer messaging suite, more turnkey and higher priced than an open platform.

Botpress wins on control and openness for developers who want to shape agent logic and place LLM reasoning where they choose. Landbot leans toward no-code marketing funnels, Voiceflow centers on collaborative agent design, and Intercom Fin offers a managed support path with less to build. For an engineering-led team that wants to own the agent, Botpress is the flexible option.

Limitations and getting started

Botpress has honest limits worth weighing. Usage-based pricing can climb with message and AI volume, the developer-first depth carries a learning curve for non-technical users, and getting the most from advanced controls rewards engineering time that a small team may lack. It is an open platform for building agents, not a one-click template.

Getting started

1. 1Create a Botpress account and open the studio to start a new agent.
2. 2Map your flow on the canvas, adding nodes for fixed logic and LLM nodes where you want model reasoning.
3. 3Add a knowledge base so the agent answers from your documents and content.
4. 4Connect the channels and integrations you need, then embed the web chat widget or deploy to your app.
5. 5Test the agent end to end, then review conversation data to refine the flow and knowledge base.

Start on the free tier, build one focused agent for a single use case, and watch your message and AI usage as traffic grows. That approach keeps cost in check while you learn where Botpress delivers the most value.