Nova

Posted on Apr 3

Botpress Review 2026: I Used It for 6 Months to Build AI Agents (Honest Verdict)

#botpressreview #aichatbotbuilder #nocodechatbot #botpresspricing

Last March, a client from Dubai approached me with what seemed like a simple request. They wanted a customer service bot for their e-commerce store that could handle order inquiries, process returns, and escalate complex issues to humans. Simple enough, right?

Photo by Victor Carvalho via Unsplash

I'd been coding chatbots from scratch for months, wrestling with APIs and spending weeks on features that should take days. That's when I stumbled across Botpress during a late-night search for "visual chatbot builder that doesn't suck."

Six months and twelve client projects later, I can tell you exactly what this tool does well and where it falls flat on its face.

What Is Botpress?

Think of Botpress as the WordPress of AI chatbots. It's a visual platform where you drag and drop components to build conversational AI agents without writing code.

Instead of typing lines of code, you create flows using boxes and arrows. Each box represents an action like "ask a question," "save user data," or "call an API." You connect these boxes with lines to create conversation paths.

The platform handles all the technical stuff in the background. Natural language processing (understanding what users mean), database connections, integrations with websites and messaging apps. You focus on designing the conversation flow.

Setting Up My First Bot

I signed up on a Thursday evening, expecting the usual hour-long setup nightmare. Surprisingly, I had a basic bot running in 37 minutes.

Here's exactly what happened:

First, I clicked "Start Building" on their homepage. No lengthy signup forms, just email and password. They sent a verification email that actually arrived in my inbox, not spam folder.

The dashboard opened with a clean interface. On the left: "Create New Bot" button. I clicked it, chose "Customer Service" from their templates, and named it "DubaiBot."

The visual flow editor loaded. It looks like a flowchart tool with a sidebar of components. I saw nodes for "Text," "Choice," "Code," and "API Call." The template already had a basic flow: greeting, menu options, and responses.

I clicked on the "Greeting" node. A panel opened on the right with a text box containing "Hello! How can I help you?" I changed it to "Welcome to our store! I'm here to help with your orders."

Testing was one click away. The "Test" button in the top right opened a chat window. I typed "Hi" and my custom greeting appeared. Magic.

Connecting it to their website took longer than building the bot. I had to copy a JavaScript snippet and add it to their site header. The live chat widget appeared in the bottom right corner within minutes.

What I Built With It

That first Dubai project became my Botpress testing ground. The client needed their bot to:

Check order status by asking for order numbers
Handle return requests with photo uploads
Route complex issues to human agents
Work in both English and Arabic

The order status feature impressed me most. I used Botpress's "API Call" node to connect with their e-commerce platform. When customers entered order numbers, the bot fetched real data and displayed shipping status, tracking numbers, and delivery dates.

Setting this up was surprisingly straightforward. I added an API Call node, entered their order lookup endpoint URL, and mapped the order number variable to the API parameter. Botpress handled the JSON parsing automatically.

Return requests were trickier. The bot needed to collect customer details, photos of damaged items, and reasons for returns. I used multiple "Capture" nodes in sequence, each collecting one piece of information.

The photo upload stumped me initially. Botpress supports file uploads, but I couldn't find clear documentation. After digging through their community forum, I discovered you need to enable "File Upload" in the channel settings first. Once enabled, users could attach images directly in chat.

For Arabic support, I used their translation feature. You can add multiple languages in the bot settings, then provide translations for each text response. The bot detects user language automatically, though it's not perfect with mixed languages.

The results? Their customer service tickets dropped by 60% in the first month. The bot handled 847 order inquiries and 203 return requests without human intervention. Response time went from hours to seconds.

What Surprised Me (Good and Bad)

The biggest surprise was how well it handled complex conversation flows. I expected the visual editor to become messy with complicated logic. Instead, I could create nested conditions, loops, and branching conversations that remained readable.

For example, I built a lead qualification bot that asked different questions based on company size, industry, and budget. The flow had 23 nodes with multiple decision points, but I could follow the logic easily months later.

The AI understanding impressed me too. Unlike basic keyword-matching bots, Botpress uses natural language processing to understand intent. Users could say "I want my money back," "Can I return this?", or "This product is defective" and the bot would route them to the returns flow.

But here's what frustrated me:

The documentation is scattered and often outdated. I spent hours searching for how to implement custom webhooks, only to find the answer buried in a community forum post. Official docs showed old screenshots that didn't match the current interface.

Version control is practically non-existent. When I accidentally deleted a node, there was no easy undo. I had to rebuild parts of flows from memory. For a platform targeting non-coders, this is unacceptable.

The mobile experience needs work. Bots look great on desktop but feel cramped on phones. Button layouts break on smaller screens, and the chat interface sometimes covers important content.

Integration limitations caught me off guard too. While they support popular platforms like Facebook Messenger and WhatsApp, connecting to custom CRMs requires coding. The "no-code" promise breaks down when you need specific integrations.

Pricing Breakdown

Botpress uses a conversation-based pricing model. Here's what each tier actually includes:

Community (Free):

5,000 messages per month
1 bot
Basic integrations (website widget, Telegram)
Community support only

This is perfect for testing or very small projects. I used it for initial client demos before upgrading.

Pro ($15/month):

10,000 messages per month
Unlimited bots
All integrations (WhatsApp, Messenger, Slack)
Email support
Custom branding removal

Most of my small business clients fit this tier. 10,000 messages sounds like a lot until you realize each user interaction counts as multiple messages.

Team ($50/month):

25,000 messages per month
Team collaboration features
Advanced analytics
Priority support
Single sign-on

I upgraded to this for larger projects. The analytics dashboard shows conversation patterns, user drop-off points, and bot performance metrics that help optimize flows.

Enterprise (Custom pricing):

Unlimited messages
On-premise deployment
Custom integrations
Dedicated support

I quoted one enterprise client who needed this tier. Pricing started at $500/month but could go much higher with custom features.

Here's the catch: message counting is aggressive. Each bot response, user reply, and API call counts toward your limit. A simple order status check might consume 8-10 messages. Monitor usage carefully or face overage charges.

Who Should Use This (And Who Should Not)

Perfect for:

Small business owners who need customer service bots but can't afford developers. If you run an online store, restaurant, or service business and want to automate common questions, Botpress delivers.

Non-technical marketers building lead generation bots. The visual editor makes sense if you think in terms of conversation flows rather than code logic.

Agencies like mine who need to build bots quickly for multiple clients. The template system and reusable components speed up development significantly.

Skip it if:

You need heavy customization or complex integrations. While Botpress handles many use cases, it struggles with unique requirements that don't fit their node system.

You're building voice-first experiences. The platform focuses on text-based chat. Voice capabilities exist but feel like an afterthought.

You need enterprise-grade security from day one. While they offer security features, the implementation feels less mature than established enterprise platforms.

You prefer coding everything from scratch. If you enjoy full control over your bot's behavior and don't mind spending weeks on basic features, stick with custom development.

My Honest Verdict After Real Projects

After six months and twelve client projects, Botpress sits in the sweet spot between too simple and too complex.

It's not the easiest chatbot builder I've used. That honor goes to Chatfuel for basic Facebook Messenger bots. But it's far more powerful than simple template-based tools.

It's not the most feature-rich platform either. Microsoft Bot Framework offers deeper customization and enterprise features. But it requires significant technical knowledge.

Botpress gives you 80% of what most businesses need without requiring a computer science degree. The visual editor makes sense once you understand the logic flow concept. Integration options cover most common use cases.

The pricing is reasonable compared to enterprise solutions, though message counting can surprise you. Documentation needs improvement, but the community forum usually has answers.

My biggest concern is long-term scalability. As bots become more complex, the visual editor becomes harder to manage. I've started hitting limitations on larger projects that would require custom coding anyway.

For most small to medium businesses, these limitations won't matter. Botpress will handle their customer service, lead qualification, and basic automation needs effectively.

Alternatives Worth Considering

ManyChat works better for simple marketing bots on Facebook and Instagram. The interface is more beginner-friendly, and pricing is more predictable. But it lacks Botpress's flexibility for complex conversations.

Dialogflow offers superior natural language processing and Google's AI capabilities. However, it requires more technical knowledge to set up and lacks Botpress's visual flow editor.

Landbot excels at creating engaging, conversational landing pages and surveys. The user experience is more polished than Botpress, but it's limited to web-based interactions.

Conclusion

Botpress earned its place in my freelancer toolkit. It's not perfect, but it solves the real problem of building functional AI agents without months of custom development.

If you're a business owner who needs a customer service bot next month, not next year, Botpress delivers. The learning curve is manageable, the results are professional, and the time savings are significant.

Just go in with realistic expectations. This is a powerful tool that still requires thinking through conversation design and user experience. The "no-code" label doesn't mean "no-thinking."

For my agency, it's become the default choice for 70% of chatbot projects. The remaining 30% need custom solutions that no visual builder can provide.

How long does it take to learn Botpress?Expect 2-3 days to understand the basics and build simple bots. Creating complex flows with API integrations takes 1-2 weeks of practice. I spent about 20 hours total before feeling comfortable with most features.

Can I export my bot if I want to switch platforms?Botpress lets you export bot configurations as JSON files, but these won't work directly on other platforms. You'll need to rebuild conversation flows manually. This is common across most chatbot builders unfortunately.

Does it work well for non-English languages?Mixed results. The platform supports multiple languages, but AI understanding works best in English. I've had good success with Spanish and French, decent results with Arabic, but struggled with languages like Hindi or Mandarin.

What happens if I exceed my message limit?Your bot continues working, but you'll be charged overage fees. It's $2 per 1,000 additional messages on Pro and Team plans. The platform sends warnings at 80% and 95% usage, giving you time to upgrade.

Can I integrate it with my existing CRM or database?Yes, through API calls and webhooks. Popular CRMs like HubSpot and Salesforce work well. Custom databases require some technical setup, but it's manageable if you understand basic API concepts. The documentation for integrations could be better though.

DEV Community