Botpress | The Complete AI Agent Platform

The Complete
AI Agent Platform

Botpress is an all-in-one platform for building AI agents powered by the latest LLMs.

Build, deploy, and monitor AI agents across channels, tools, and data.

Botpress provides the core infrastructure needed to build and run AI agents in production.

The core of every Botpress agent is LLMz, a custom inference engine. It’s responsible for coordinating the agent’s behavior: interpreting instructions, managing memory, choosing the right tools, executing code, and returning a usable response.

Unlike standard tool-calling frameworks that rely on rigid prompts and external orchestration, the Botpress custom inference engine runs everything internally. It can handle complex, multi-step logic, execute JavaScript in a safe sandbox, and generate structured responses without extra configuration.

Engine

##Identity
You are a customer support agent. Your goal is to route the user to the correct path based on what they're looking to accomplish in today's conversation.

Botpress uses a fully isolated runtime architecture: every deployed agent runs its own self-contained environment.

They're

versioned

,

durable

,

and

compatible

with

future

platform

changes,

so

what

you

build

today

will

still

work

tomorrow.

Developers

By developers, for developers.

Custom code

Inject custom code into an agent’s lifecycle events.

Inspect

Maintain full observability over an agent's actions and executions.

API

Access the API endpoints you need to build and manage agents.


const options = {
  method: 'POST',
  headers: {
  'x-user-key':'<x-user-key>',
  'Content-Type': 'application/json'
  },
  body: '{
  "payload":{"audioUrl":"<string>", "type":"audio"},
  "conversationId":"<string>"
  }'
};

fetch('https://chat.botpress.cloud/{webhookUrl}/messages', options)
  .then(response => response.json())
  .then(response => console.log(response))
  .catch(err => console.error(err));

const options = {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: '{
  "name":"<string>",
  "factor":1,"frozen":true,
  "schema":{},
  "tags":{},
  "isComputeEnabled":true}'
};

fetch('https://api.botpress.cloud/v1/tables', options)
  .then(response => response.json())
  .then(response => console.log(response))
  .catch(err => console.error(err));
  

const options = {
  method: 'POST', headers: {'Content-Type': 'application/json'},
  body: '{
  "states":{},"events":{},"recurringEvents":{},
  "subscriptions":{"events":{}},"actions":{},
  "configuration":{"data":{},"schema":{}},
  "user":{"tags":{}},"conversation":{"tags":{}},
  "message":{"tags":{}},"tags":{},"code":"<string>","name":"<string>",
  "medias":[{"url":"<string>","name":"<string>"}],
  "url":"<string>","dev":true}'
};

fetch('https://api.botpress.cloud/v1/admin/bots', options)
  .then(response => response.json())
  .then(response => console.log(response))
  .catch(err => console.error(err));

LLM tokens processed per hour

42,483,460

An illustration of books, a plant, and a laptop on a table.