Botpress

Botpress is an AI agent platform for building, deploying, integrating, and monitoring conversational AI agents. It combines a visual agent-building environment with developer tools, knowledge bases, APIs, integrations, analytics, and deployment capabilities, allowing organizations to create AI agents for customer support, automation, and other business workflows.

Botpress began as an open-source conversational AI project and was founded in 2017. The company initially focused on giving developers tools to build conversational applications without requiring large data-science or machine-learning teams. It has since evolved into a cloud-based AI agent platform designed around large language models and autonomous agents.

The current platform is Botpress Cloud. Botpress describes it as an all-in-one environment for building AI agents powered by modern LLMs. It handles the infrastructure required to build and operate agents in production, so users do not need to manage their own Botpress servers.

Botpress Studio provides a visual drag-and-drop interface for creating agents and conversation workflows. Users can design workflows, add nodes and cards, configure instructions, connect knowledge sources, and test their agents through an emulator. The platform can be used with little or no code, while developers can add custom code when more advanced behavior is required.

For developers who prefer a code-first approach, Botpress provides the Botpress ADK, a TypeScript library for building AI agents from code. The platform also provides APIs for managing bots, conversations, files, tables, integrations, analytics, and runtime operations.

A major part of Botpress is its Knowledge Base system. Agents can use documents, websites, text, tables, and other sources as knowledge. Files can be indexed for semantic search and used in retrieval-augmented generation workflows, allowing agents to answer questions using an organization’s own information.

Botpress also includes specialized AI agents that extend an agent’s capabilities. These include a Knowledge Agent for retrieving information, a Translator Agent for multilingual conversations, a Summary Agent for generating conversation summaries, a Personality Agent, a Policy Agent, a Vision Agent, an Analytics Agent, and a Human-in-the-Loop Agent.

The Translator Agent can automatically identify the user’s language and translate messages, allowing a bot to communicate across multiple languages. Botpress exposes the user’s detected language as an ISO 639-1 language code.

The platform also supports vision and multimodal capabilities. The Vision Agent can add image recognition and visual processing capabilities to an agent, enabling workflows that can work with visual information in addition to text.

Botpress supports stateful and persistent conversations, allowing agents to retain context across steps. This is useful for multi-step customer-support conversations and workflows where the agent needs to remember information during an interaction.

Another important capability is custom actions and code execution. Developers can inject custom code into an agent’s lifecycle and connect the agent to external APIs and services. This allows Botpress agents to perform actions rather than simply return text responses.

Botpress includes Tables for structured data. Tables can store information such as customer records, workflow data, or other structured information that an agent can access. The platform’s Tables API allows developers to create, update, search, and manage table data programmatically.

The platform also provides Files and Vector Database storage. Files can be uploaded and indexed for knowledge-base searches, while vector storage supports semantic retrieval. The Files API can manage documents, images, videos, and other files used by bots and integrations.

Botpress has a Botpress Hub, a marketplace for integrations, workflows, and plugins. Users can search for integrations, install them into their bots, and configure the necessary credentials or settings. Developers can also build their own integrations using the Botpress SDK and publish them to the Hub.

The platform supports major communication channels including WhatsApp, Telegram, and Messenger, while the wider integration ecosystem allows bots to connect with external services and business systems.

For customer-support teams, Botpress provides Botpress Desk, a helpdesk designed around AI agents. It supports monitoring conversations, managing escalations, and allowing human agents to intervene when necessary.

Botpress also provides a documented Zendesk integration. The integration can create and manage support tickets and allow a Zendesk agent to take over a conversation that began with a Botpress agent.

The platform offers Webchat, allowing an agent to be embedded directly into a website. Developers can customize the appearance and behavior of the web chat experience and deploy the agent using Botpress’s web integration.

Botpress also provides a Cloud API consisting of several APIs, including Chat, Admin, Files, Tables, and Runtime APIs. These APIs allow developers to interact with agents programmatically, manage workspaces, retrieve analytics, handle files, manipulate structured data, and manage conversations.

The official TypeScript client can be installed through npm, pnpm, or yarn, making it possible for developers to interact with Botpress programmatically from TypeScript and JavaScript applications.

Botpress supports multiple LLM providers. Its current pricing documentation lists providers such as OpenAI, Anthropic, and Groq, and the platform also supports backup LLMs and bring-your-own LLM/API keys on eligible plans.

A notable feature is LLM usage management. Botpress separates its platform subscription from AI Spend. AI Spend represents the cost of LLM usage and is charged at the provider’s cost without Botpress markup. Each workspace receives a monthly AI credit, after which additional AI usage is charged according to consumption.

Botpress also provides analytics and observability. Developers can inspect logs, errors, bot activity, conversation data, and performance information. Higher plans add custom analytics dashboards and additional reporting capabilities.

The platform supports human handoff, allowing an AI agent to transfer a conversation to a human when required. This is particularly useful for customer support workflows where the AI can handle routine questions while more complex or sensitive cases are escalated.

For teams, Botpress provides collaboration capabilities. Higher-tier plans include role-based access control, real-time collaboration, workspace permissions, staging and production environments, and custom analytics.

Botpress is now primarily a cloud platform. Its older Botpress v12 and self-hosted versions have officially been sunset and are no longer available for new deployments. New users are directed to Botpress Cloud.

The current pricing structure includes Pay-as-you-go, Plus, Team, Managed, and Enterprise. The Pay-as-you-go plan starts at $0 per month, while Plus, Team, and Managed are paid plans and Enterprise uses custom pricing. AI Spend is charged separately based on LLM consumption.

The Pay-as-you-go plan includes a visual building studio, a $5 monthly AI credit, community support, and a base allowance of 500 incoming messages/events per month. The Plus plan is listed at $89/month, with an annual-billing price of $79/month, and adds capabilities such as human handoff, conversation insights, watermark removal, proactive chat, visual knowledge-base indexing, and live-chat technical support.

The Team plan is listed at $495/month, or $445/month with annual billing. It adds role-based access control, real-time collaboration, custom analytics, and advanced support. The Managed plan provides a more hands-off service where Botpress experts build, deploy, maintain, and optimize AI agents for customers.

Enterprise pricing is customized and includes additional capabilities such as white-glove onboarding, custom workspace limits, and a dedicated support manager.

Botpress is best suited to developers, customer-support teams, businesses, startups, and enterprises that need customizable AI agents capable of handling conversations, accessing business information, triggering workflows, connecting to external systems, and escalating cases to humans.

Its combination of visual development, code-level customization, knowledge bases, integrations, APIs, multilingual support, analytics, and customer-support tooling makes it a broad AI-agent platform rather than a simple chatbot builder.

Pricing

Botpress uses a freemium, subscription-based pricing structure with usage-based AI Spend. The current official pricing page lists Pay-as-you-go at $0/month, Plus at $89/month, Team at $495/month, Managed at $1,245/month, and Enterprise at custom pricing. Annual billing currently shows discounted rates of $79 for Plus, $445 for Team, and $995 introductory pricing for Managed.

Pricing Model

Paid

Pricing Note

Botpress has a free Pay-as-you-go plan, but its overall pricing model is Paid because the platform offers multiple paid subscriptions. AI Spend is billed separately at provider cost without markup.

Review

Botpress is a capable platform for organizations that want to build AI agents without giving up developer-level control. The combination of its visual Studio, knowledge bases, APIs, integrations, custom code, and AI agents makes it flexible enough for both simpler chatbot projects and more complex business workflows.

Its biggest strengths are customization and connectivity. Agents can connect to external APIs and services, use organizational knowledge, work across messaging channels, and transfer conversations to human agents.

The platform also receives positive feedback for its visual interface and integration ecosystem. Current G2 reviews mention ease of setup, a clean interface, strong community support, and a wide selection of integrations.

The main consideration is the learning curve. Botpress provides a large number of features, APIs, integrations, workflows, variables, knowledge tools, and customization options, so users seeking only a very simple chatbot may find the platform more extensive than necessary.

The pricing structure also requires attention because the subscription fee is separate from AI Spend. Organizations with significant LLM usage should monitor their AI consumption in addition to their platform subscription.