P
PNPBrain
FeaturesHow it worksPricing
Log inGet started free
Powered by local Ollama and LangGraph.js

Your customers deserve an intelligent assistant

PNPBrain embeds a RAG-powered AI assistant on your website, app, or channel in minutes. It learns from your knowledge base, remembers customers, and handles questions 24/7.

Why choose PNPBrain: most tools stop at answers. PNPBrain answers, takes action, and helps convert visitors into customers.

Start for freeSee how it works
P
PNPBrain Assistant
Online

Hi! How can I help you today?

What are your business hours?

Monday-Friday 9AM-6PM and Saturday 10AM-4PM.

Type a message...

Everything your AI assistant needs

A full-stack, production-ready platform, not just a chatbot wrapper.

🧠

RAG-Powered Knowledge Base

Upload documents, paste URLs or run Firecrawl to crawl your site. PNPBrain chunks, embeds, and retrieves the right context for every answer.

💾

Long-Term Customer Memory

PNPBrain stores conversation history, keeps useful context across sessions, and lets customers revisit past chats without starting over.

🔌

One-Line Embed

Drop a single <script> tag on any HTML page or install the WordPress plugin. The widget is self-contained and < 50 KB.

🏃

Runs Locally with Ollama

No cloud API keys required in development. Pull `llama3.1:8b` and start building. Switch to GPT-4 or Claude with one env var.

🔒

Domain-Scoped Web Scraping

Give Firecrawl permission to index only your approved domains. The agent can never scrape arbitrary URLs at runtime.

📊

Admin Dashboard

Manage the knowledge base, review conversations, monitor crawl jobs, and tweak the widget appearance — all from one place.

Interactive examples

Click through real integration patterns

Pick an integration path, switch languages, and copy snippets. Every example below follows the actual endpoints and payload contracts in this codebase.

Integration paths
Native WordPress path (plugin + shortcode)
WordPress Plugin Embed

Matches the WordPress plugin in this repo: configure settings once, then use shortcode or auto-inject.

[pnpbrain_widget]
Settings -> PNpbrain Widget
auto_inject
PNPBRAIN_CONFIG
[pnpbrain_widget
  public_token="PUBLIC_CHAT_TOKEN"
  backend_url="https://api.pnpbrain.com"
  bot_name="Support"
  primary_color="#6366f1"
  welcome_message="Hi! How can I help you today?"
]
1) Choose transport

Web SDK tries WebSocket first (`/ws/agent`) and falls back to SSE (`POST /api/agent/chat`) automatically.

2) Persist context

Keep one `threadId` per end-user or per channel conversation ID so memory and context continue correctly.

3) Keep secrets safe

Use `publicToken` in browser/public surfaces. Keep `x-api-key` strictly on server-side integrations and channel bridges.

How it works

Four steps from signup to live AI assistant.

01
Sign up & connect your site

Create an account, name your assistant, and paste your website URL. PNPBrain is ready in under 2 minutes.

02
Build your knowledge base

Upload PDFs, paste content, or let Firecrawl automatically crawl approved pages. Content is chunked and embedded automatically.

03
Embed the widget

Copy one <script> tag into your HTML, or install the WordPress plugin. The widget appears instantly on your site.

04
Watch it work

Customers get accurate, instant answers 24/7. You monitor conversations and refine the knowledge base from the admin dashboard.

Why teams choose PNPBrain over alternatives

Not another chatbot widget. PNPBrain combines accurate answers, real actions, memory, and integrations so your site can convert and support customers 24/7.

Most chatbots only reply. PNPBrain gets work done.

Book meetings, qualify leads, look up orders, and escalate support issues with full context from the same conversation.

Context memory that improves customer experience

PNPBrain remembers key details across sessions, so your customers do not have to repeat themselves every time they return.

Fast setup on any site

Go live with a WordPress plugin, script embed, React component, or API without replatforming your existing stack.

Built for trust, privacy, and real outcomes

Privacy-first architecture, optional on-prem deployment, and measurable impact with strong query deflection and conversion gains.

Typical chatbot

Answers basic FAQ and stops there

PNPBrain

Answers accurately and completes real business actions end-to-end

Typical chatbot

Loses context between sessions

PNPBrain

Uses long-term memory for personalized follow-ups

Simple, transparent pricing

Start free. Upgrade when you're ready.

Freemium

Best for trying PNPBrain before committing.

$0/ month

1 website

200 conversations / month

5 MB knowledge base

PNPBRAIN subdomain hosting

Community support

Get started free
Lite

Simple monthly plan for early production use.

$19/ month

3 websites

2,000 conversations / month

100 MB knowledge base

Custom domain support

Threaded chat history

Email support

Start Lite
Most popular
Basic

For growing teams that need predictable volume.

$49/ month

10 websites

10,000 conversations / month

500 MB knowledge base

Custom domain support

Priority email support

Start Basic
Pro

High-volume operations with faster response support.

$149/ month

30 websites

50,000 conversations / month

2 GB knowledge base

Advanced analytics

Priority support

Start Pro
Custom

Tailored package with direct onboarding support.

Custom

Unlimited websites

High volume conversation limits

Custom LLM provider

Optional self-hosted deployment

Security/compliance add-ons

Dedicated support

Contact support