Quick verdict: different categories, overlapping search bucket
Botpress and ChatRaj both show up when you search for "AI chatbot platform", but they are not the same kind of product. Treating them as direct substitutes leads to the wrong call about half the time, so the first job of this page is to separate the categories before anyone looks at price.
Botpress is an open-core conversational AI platform. Its center of gravity is a visual flow builder where you drag nodes onto a canvas, wire them together, and design exactly how a conversation should branch. Around that builder sit a workspace, an event-driven runtime, a marketplace of integrations and skills (the Hub), and deployment connectors to a long list of messaging channels. The current main repository on GitHub is MIT licensed, which makes it genuinely fork-friendly. The legacy v12 line is dual-licensed under AGPLv3 and a Botpress proprietary license, which matters for anyone still self-hosting that generation.
ChatRaj is a hosted SaaS chatbot for websites. There is no canvas, no node graph, and no flow builder. You paste a URL, ChatRaj crawls your site, builds a hybrid (semantic plus keyword) retrieval index over the content, and serves answers through an embeddable widget. Pricing is flat per month for a fixed message quota: Pro at twenty-nine dollars for ten thousand messages, Growth at ninety-nine dollars for fifty thousand. There is no self-host option.
If you imagine the products as physical shapes, Botpress is a workshop and ChatRaj is an appliance. The workshop gives you more freedom and more responsibility. The appliance is faster to plug in and works the same on every site, but you do not get to redesign it. Most teams need exactly one of these, not both.
Botpress strengths: open source, visual flow builder, multi-channel
The case for Botpress is not subtle. Three structural advantages keep showing up in real evaluations.
First, the license. The current Botpress main repo on GitHub is MIT licensed and the codebase is actively maintained. You can fork it, run it inside your own VPC, audit every line, and avoid vendor lock-in entirely. The v12 line uses dual AGPLv3 plus proprietary licensing, which is a different posture but still fundamentally an open codebase. For regulated industries, government, and large enterprises with a hard requirement that the LLM application stack be inspectable and self-hostable, this matters more than any feature comparison.
Second, the visual flow builder. Botpress's canvas is one of the most mature drag-and-drop bot builders on the market and has been refined since the v11 days. You drop nodes for messages, intents, conditions, API calls, sub-flows, and (in the modern version) autonomous LLM nodes that can reason within a structured flow. Complex branching dialogs, hand-off rules, escalation paths, A/B variants, and stateful multi-turn collection (think "ask for name, then email, then issue type, then route to the right team") are exactly what this kind of interface is good at. ChatRaj has nothing equivalent because ChatRaj is not trying to do that job.
Third, the channels. Botpress deploys from a single bot to web chat, WhatsApp, Facebook Messenger, Instagram DM, Telegram, Slack, Microsoft Teams, Viber, and several more, plus custom channels through its event API. If you want one bot that answers in WhatsApp during the day and Slack inside your company at night, Botpress is set up for that out of the box. ChatRaj currently ships a website widget only. WhatsApp, Slack, and Teams are not on the ChatRaj surface, full stop.
A fourth, softer strength: the Hub. Community-shared skills, integrations, and bot templates mean a Botpress builder rarely starts from a blank canvas. That ecosystem effect compounds over time and is a real reason to pick Botpress for a long-lived programme of bots, not a single quick widget.
ChatRaj strengths: hosted, RAG-from-content default, flat pricing
The case for ChatRaj is also not subtle, but it is narrower.
ChatRaj is built around one assumption: most people who want an "AI chatbot for my website" actually want a thing that reads their existing content and answers visitor questions about it. They do not want to design a conversation tree. They do not want to wire conditions. They want a widget that knows their pricing page, their FAQ, their docs, and their blog, and responds in coherent prose.
That assumption shapes the whole product. The default flow is paste URL, wait for crawl, paste the widget script, done. Hybrid retrieval (BM25-style keyword plus vector semantic search) runs out of the box, so the first answers a visitor gets are already grounded in your content rather than the model's pre-training. There is no flow to design because there is no flow. The bot reads, retrieves, answers, and cites.
Hosted means zero ops. You do not run a database. You do not patch a runtime. You do not pay for a vector database, an LLM API key, a Redis instance, and a CDN. You pay one monthly fee. The hosting cost, model cost, and embedding cost all collapse into that fee. For a small marketing team or a solo founder, this is the difference between shipping a bot in twenty minutes and shipping nothing.
Pricing is flat and predictable. Pro is twenty-nine dollars per month for ten thousand messages, Growth is ninety-nine dollars per month for fifty thousand messages, and there are no per-conversation overages, no AI credit balances to top up, and no per-channel add-ons. If you go over a quota you upgrade a tier; there is no surprise invoice.
Speed of setup is the other piece. A ChatRaj bot from sign-up to "answering questions on a real production page" usually takes well under an hour, often closer to fifteen minutes. A Botpress bot to roughly the same level of usefulness typically takes a day or more of canvas work, even with the Hub templates as a head start, because the canvas presumes you are designing something custom.
When Botpress is the right call
Several scenarios point to Botpress, not ChatRaj. Take them seriously; if any of these describe you, Botpress is the better fit and this page will not try to talk you out of it.
You need a complex, branching, deterministic flow. A loan application bot. A pre-authorisation triage bot for an insurance line. A booking flow with three conditional paths based on customer type. These are jobs where the conversation logic matters more than retrieval, and a visual builder is the right tool.
You need WhatsApp, Slack, Telegram, Teams, Messenger, or Instagram as a primary channel. ChatRaj does not currently deploy to any of those. Botpress does.
You have a hard self-host or on-premises requirement. Regulated industries, government, defence-adjacent work, or just a security team that has decided no customer data leaves the VPC. Botpress's MIT main repo and its v12 OSS lineage are designed for this. ChatRaj is hosted only.
You have engineering capacity and want to own the stack long term. Forking Botpress, building custom integrations, plugging it into your CI, running it in your Kubernetes cluster, this is a sensible programme if you have the team and the appetite. ChatRaj is intentionally the opposite trade.
You are building a programme of dozens of bots across business units. The canvas, the Hub, the workspace model, and the team features in Botpress's higher tiers scale to that shape. ChatRaj is built for "one bot per website".
When ChatRaj is the right call
The picture flips when the job is narrower.
You have a website. You want a widget on it. The widget should answer questions from your existing content. You do not want to design a flow, run a vector database, or pay four invoices. You want a flat predictable monthly bill. You want it live this afternoon. That sentence is the ChatRaj target customer almost word for word, and for that customer ChatRaj is faster, cheaper, and less work than Botpress by a wide margin.
Concretely, ChatRaj wins for SaaS marketing sites with a docs subdomain, e-commerce stores that want pre-sales Q and A on product pages, service businesses (clinics, agencies, contractors) that want after-hours intake, and solo founders or small content teams who need a chatbot but have no spare engineering hours.
Pricing math at SMB and growth scale
Pricing comparisons between these two products are tricky because the units are different. Botpress sells seats, AI spend, storage, and channel tiers; ChatRaj sells flat message quotas. Two worked examples:
Small business, around ten thousand visitor messages per month. ChatRaj Pro is twenty-nine dollars per month flat. Botpress Plus is eighty-nine dollars per month and bundles three seats, unlimited AI agents, WhatsApp, whitelabel webchat, one gigabyte of vector storage, and human handoff. On top of the Plus base, Botpress charges separately for AI spend (the underlying model and embedding costs the platform passes through). At ten thousand messages with a modern model, AI spend can sit anywhere from a handful of dollars to several tens of dollars depending on model choice and prompt size. Total Botpress cost at this volume is roughly ninety to one hundred and thirty dollars per month, three to four times ChatRaj Pro. But for that premium you get the visual builder, multi-channel, and whitelabel.
Growth team, around fifty thousand messages per month. ChatRaj Growth is ninety-nine dollars per month flat. Botpress at this volume is typically on the Team tier (around four hundred and ninety-five dollars per month) plus AI spend, plus any extra storage or channels needed. The gap widens. Botpress is meaningfully more expensive at growth scale unless you are using the things ChatRaj does not have (channels, advanced flows, multi-bot workspaces).
The honest read: at the same message volume Botpress almost always costs more. That is not a knock; you get more product. The question is whether you need more product.
Migration path from Botpress to ChatRaj
A migration only makes sense if the original Botpress bot was doing the job ChatRaj does, which is "answering questions from website content". Bots built around flows or non-web channels should not be migrated; they should stay on Botpress. The migration below assumes a content-answering bot that has drifted into being more trouble than it is worth.
The five-step process is in the section below. The short version: export your knowledge base content, point ChatRaj at the same source URLs, validate that the answers match, swap the widget script, and decommission the Botpress workspace once the new bot has run clean for a week.
What ChatRaj does NOT have that Botpress does (rows where Botpress wins)
This section exists because most self-comparison content lies by omission. ChatRaj does not have the following, and if you need any of them, Botpress is the answer:
- A visual flow builder. There is no canvas in ChatRaj. None.
- Multi-channel deployment. ChatRaj is a website widget. No WhatsApp, no Slack, no Teams, no Telegram, no Discord, no Messenger.
- Self-hosting. ChatRaj is hosted SaaS only. You cannot run it on your own infrastructure.
- An open-source codebase you can fork. ChatRaj is proprietary.
- A community Hub of shared skills and templates.
- Whitelabel webchat at the entry tier. Botpress Plus includes whitelabel at eighty-nine dollars per month; ChatRaj reserves whitelabel for higher tiers and enterprise.
- Complex branching dialog design with deterministic conditional logic across many steps. Possible to approximate with prompt instructions in ChatRaj, but not the same product.
If two or more of those are real requirements for your team, stop reading and go set up a Botpress workspace. That is the honest call.
If none of them are requirements, ChatRaj is faster, cheaper, and less work, and the rest of the page applies.