Quick Answer: Which AI chatbot builders have a free plan?
Several well-known AI chatbot builders have offered a free plan, including Chatbase, Voiceflow, Botpress, Tidio, ManyChat, Landbot, and Google’s Dialogflow. Each one gives you a way to build and test a bot without paying upfront, but the free tiers differ sharply in what they limit.
Some free plans cap the number of AI messages or credits per month. Others cap contacts, live chats, published bots, or seats. A few restrict branding removal or API access to paid tiers, so the free version stays visibly “free.”
Before committing, check the capped metric, the overage rules, branding requirements, publishing limits, and how easily you can export your work later. Free tiers in this market change often, so always confirm the current terms on each vendor’s own pricing page before you build.
Why free tiers are trickier than they look

A free chatbot plan sounds like a gift until you hit the wall. The wall is usually invisible at sign-up. You build a beautiful flow, connect your knowledge base, test a dozen conversations, and then a banner appears: you have used 30 of your 30 free messages. That is the moment most people learn what “free” actually means.
The chatbot market has a naming problem. “Free plan” can mean five very different things:
- A usage cap on AI messages, credits, or interactions per month
- A contact cap, common in messaging-automation tools tied to social platforms
- A chat cap, where only a set number of conversations can happen
- A seat cap, limiting how many teammates can log in
- A feature cap, where publishing, branding removal, or API access sits behind a paywall
None of these is dishonest. They are just different. The mistake is comparing two “free” plans without checking which metric each one limits. A tool that gives you 1,000 free messages but forces your bot to wear its logo may serve you worse than one with 100 messages and full white-labeling.
There is a second trap: stale listicles. Many “best free chatbot” roundups are affiliate-driven and quietly out of date. Free tiers get grandfathered, shrunk, or discontinued with little warning. The only reliable check is the vendor’s own pricing page, read on the day you sign up.
The main builders with a free entry point
Here is a plain-language look at the tools that have historically offered a free tier. Treat every line as “as last known” rather than gospel, because terms shift.
Chatbase
Chatbase built its reputation on turning a document set or website into a support bot fast. It has offered a free tier with a limited number of message credits and a single chatbot. That is enough to test retrieval quality on your own content. The catch is credits: a busy support widget can burn through a small allowance in days, so the free plan is really an evaluation tier rather than a production tier.
Voiceflow
Voiceflow has offered a free sandbox for designing conversational assistants, with paid plans for publishing and heavier usage. It is the most design-forward option on this list. Teams that care about conversation architecture, variables, and multi-step flows tend to like it. The free sandbox is generous for building; the friction arrives when you want to ship to a real channel at volume.
Botpress
Botpress has offered a free plan with a monthly quota of AI interactions, plus paid tiers above it. It sits closer to a developer platform than a plug-and-play widget, which is good news if you want control and less good if you want a bot live in twenty minutes. If you are weighing developer-leaning tools against friendlier ones, a side-by-side like Landbot vs tidio 2027 which no code chatbot builder fits helps frame the trade-off.
Tidio (Lyro AI)
Tidio has offered a free plan for its chatbot and live-chat product, with limits on AI conversations and paid tiers above. Its strength is the blend of automation and human handoff, which suits small support teams. The free tier is genuinely usable for a low-traffic site, but the AI conversation allowance is the number to watch.
ManyChat
ManyChat has offered a free tier for messaging automation, with paid plans for higher contact counts and advanced features. It is the go-to for Instagram and Messenger flows rather than website widgets. Because pricing tracks contacts, a viral post can push you over the line quickly. For a broader comparison of this style of tool, see Chatfuel vs manychat 2027 which chatbot builder fits.
Landbot
Landbot has offered a free Starter plan with a monthly cap on chats and paid plans above it. It excels at conversational forms and lead capture, where the bot feels like a chat rather than a menu. The chat cap is the constraint: fine for a landing page test, tight for a always-on support widget.
Custom GPTs and Dialogflow
Building a custom GPT has historically been available to ChatGPT users, but which plan tier grants access to the builder and to GPTs has changed more than once, so treat this as the most volatile option on the list. Google’s Dialogflow has offered a free tier with usage quotas and pay-as-you-go pricing beyond, making it a solid choice for developers who want an API-first agent with clear quota math.
Free plan comparison at a glance
| Builder | What the free tier has historically capped | Best fit |
|---|---|---|
| Chatbase | Message credits, one bot | Fast support bot from docs |
| Voiceflow | Building in a sandbox; publishing is paid | Conversation design |
| Botpress | Monthly AI interactions | Developer control |
| Tidio | AI conversations | Small support teams |
| ManyChat | Contacts | Social messaging flows |
| Landbot | Monthly chats | Lead capture forms |
| Custom GPTs | Plan-dependent access | Personal assistants |
| Dialogflow | Usage quotas | API-first agents |
Read that table as a starting map, not a price list. The capped metric tells you where the ceiling is, and the ceiling tells you whether the tool fits your traffic.
How to test a free plan without wasting a weekend
The practical move is to treat the free tier as a two-hour experiment, not a long-term home. Build the smallest useful bot, throw realistic questions at it, and watch where it breaks.
- Pick your worst-case question set. Grab ten real questions from your inbox or support log, including the awkward ones. Paste them into the bot. A free tier that handles eight of ten is promising; one that handles three is not.
- Count the credits per conversation. One user question might cost one credit, or several if the bot retrieves, reasons, and rephrases. Multiply by your expected daily chats to see how fast the free allowance evaporates.
- Check the branding. Ask whether the free plan forces a “Powered by” badge. On a customer-facing widget, that badge is a real cost.
- Test the export path. Before you invest hours in flows, confirm you can export or at least document your setup. Lock-in is the hidden fee of free plans.
- Read the overage rules. Some tools pause the bot when credits run out; others keep serving and bill you. Know which one you are on.
A common mistake is building the whole bot on the free tier and only then reading the pricing page. Flip the order. Ten minutes of reading saves a rebuild.
If hallucinated answers are your bigger worry, the free plan will not solve it. Model behavior matters more than tier, and a comparison like AI chatbot hallucinations 2027 which models lie least is a better use of your time than chasing extra credits.
Where free plans quietly cost you
Free tiers differ from smaller paid plans. They change how your bot behaves in public, and that has downstream effects.
The first cost is trust. A support bot that runs out of messages mid-conversation looks broken, not thrifty. If your free allowance pauses the widget at 2 p.m. on a busy day, visitors see a dead chat window. That is worse than having no bot at all.
The second is data. Free plans sometimes limit knowledge-base size, retention, or analytics. You may not be able to see what people asked, which means you cannot improve the bot. A free tier with no conversation logs is a demo, not a tool.
The third is growth math. Contact-capped and chat-capped plans scale with your success, which is a strange incentive. The month your marketing works is the month you get an upgrade prompt. Budget for that before it happens.
The fourth is switching cost. Flows built in one builder rarely port cleanly to another. The longer you stay on a free tier, the more your setup is worth and the more painful the move becomes. Deciding early whether a tool is a keeper is cheaper than deciding late.
When paying beats staying free
The honest answer for most teams: free is for evaluation, paid is for production. The crossover usually arrives when one of three things happens.
You hit the cap during normal traffic rather than a stress test. You need branding removed because the widget sits on a customer-facing page. Or you need analytics, integrations, or API access that the free tier locks away.
At that point, compare the paid tiers the same way you compared the free ones: by the capped metric. A plan that charges by contacts punishes audience growth. A plan that charges by messages punishes engagement. A flat seat-based plan punishes nothing until you hire. Match the pricing model to your traffic pattern, not to the sticker price.
There is also a middle path many people miss: run a free tier for a low-stakes internal bot, like an FAQ helper for staff, while paying for the one customer-facing bot. Two tools, one budget line. It sounds messy, but it often beats overpaying for a single platform.
If you are still mapping the landscape, a wider ranking like 12 best AI chatbot builders 2027 no code ranking that holds gives context on which tools hold up once the free tier ends.
Common mistakes people make with free chatbot plans
The same errors show up again and again, and most are avoidable with a little skepticism.
Trusting a listicle over the pricing page. Secondary sources go stale. The vendor’s own page is the only current source of truth, and even that deserves a date stamp in your notes.
Confusing “free to build” with “free to run.” Several platforms let you design for free and charge at publish. That is a fair model, but it is not a free chatbot.
Ignoring the credit multiplier. One message is not always one credit. Retrieval, rephrasing, and tool calls can each cost. Estimate conservatively.
Skipping the branding check. A free badge on a customer widget is a marketing decision and a pricing one.
Building before checking export. Hours of flow design are worth nothing if you cannot take them with you.
Forgetting that terms change. A free tier you rely on can shrink. Keep a fallback tool in mind, even if you never use it.
Conclusion
Free plans are real, and several capable builders offer them. Chatbase, Voiceflow, Botpress, Tidio, ManyChat, Landbot, and Dialogflow have all given people a no-cost way in. The differences that matter are not the headline “free” label but the metric each one caps and the features it holds back.
The practical playbook is simple. Pick two tools, build the smallest useful bot in each, test with your ten hardest questions, and read the pricing page before you fall in love with the flow. Free tiers reward people who test fast and decide early. They punish people who build first and read later.
One last note on timing: free-plan terms in this market move constantly. Whatever you read here, confirm it on the vendor’s own pricing page on the day you sign up, and write the date next to your notes. That single habit will save you more money than any comparison table.
FAQ
Do free chatbot plans expire?
Some do. A few vendors offer time-limited trials dressed up as free plans, while others keep a permanent free tier with usage caps. Check whether the plan is a trial with an end date or a standing tier, and note what happens to your bot when the period ends.
Can I use a free chatbot plan on a commercial website?
Usually yes, with conditions. Many free tiers allow commercial use but cap messages or chats and require a visible badge. Read the terms for restrictions on client-facing deployments, and confirm whether white-labeling is a paid feature.
What happens when I hit the free plan limit?
It depends on the vendor. Some pause the bot until the next billing cycle, some keep it running and prompt you to upgrade, and some stop only the AI features while basic flows continue. Know which behavior applies before your busiest day.
Is a free plan good enough for real customer support?
For low traffic, sometimes. For anything steady, rarely. Free allowances are sized for evaluation, and a bot that dies mid-conversation damages trust more than having no bot. Use free tiers to prove the concept, then budget for a paid tier before launch.
Which free plan is best for someone with no coding skills?
Look for builders with visual flow editors and generous sandboxes. Landbot and Tidio are usually friendlier for non-technical users, while Botpress and Dialogflow assume more technical comfort. The best test is building one small flow in each and seeing which one you finish first.
Do free plans include analytics and conversation logs?
Not always. Analytics and logs are common paywall features, and without them you cannot see what customers asked or where the bot failed. If improvement matters to you, check the analytics limits on the free tier before you commit.


