Short answer: For most institutional Drupal sites in 2026, Asyntai is the option we recommend: it installs through the Header and Footer Scripts module or a one-line addition to your theme's html.html.twig, works across Drupal 7 through 10 and newer without version-specific releases, answers using your own content, replies in each visitor's language automatically, and starts free with 100 messages per month. The five alternatives below — the open-source Drupal AI module ecosystem, Zendesk, Tidio, Elfsight and LiveChatAI — represent the other routes a Drupal web team is likely to shortlist.
Drupal is not a small-business platform. It runs university faculties, ministries, city portals, hospital networks, NGOs and research institutions — sites with thousands of nodes, strict editorial governance, and a web team that answers to a procurement office rather than a personal credit card. So a listicle written for a boutique shop owner ("just paste the code and watch sales roll in!") misses what a Drupal evaluation actually looks like.
When a university or public-sector web manager assesses an AI chatbot for Drupal, the questions are different. Does it require a module deployment through our dev–stage–prod pipeline, or can it ship as configuration? What happens to it when we upgrade Drupal core next year? Where do conversation logs live, and who can we point our data-protection officer at? Can it answer a prospective student writing in Vietnamese at 3 a.m.? And crucially: who maintains this thing in year two, when the person who set it up has moved on?
This guide compares six options through exactly that lens. One of them — the contributed AI module ecosystem — is not a vendor at all, but a build-it-yourself route that deserves an honest treatment, because for some Drupal teams it is genuinely the right call.
Quick comparison table
| Option | How it reaches Drupal | Pricing model | Fits when |
|---|---|---|---|
| AsyntaiEditor's pick | JS snippet via Header and Footer Scripts module, theme template, or asset library | Free (100 msgs/mo); Starter $39/mo (2,500); Standard $139/mo (15,000); Pro $449/mo (50,000) | You want answers from your existing content, in any language, with near-zero maintenance |
| Drupal AI module ecosystem | Contributed modules installed via Composer | Open source; you pay LLM API usage plus development and maintenance time | You have in-house Drupal developers and need everything inside Drupal |
| Zendesk | Web widget script | Per-agent subscription; check the vendor's current pricing | The institution already runs Zendesk as its service desk |
| Tidio | JS script | Free tier; paid plans with separately metered AI; check the vendor's current pricing | A smaller Drupal site that wants live chat and AI bundled |
| Elfsight | Embeddable widget script | Per-widget subscription; free tier; check the vendor's current pricing | You already use Elfsight widgets and want one more from the same panel |
| LiveChatAI | JS script | Tiered plans metered by AI usage; check the vendor's current pricing | You want a content-fed AI chat and are comparing hosted options |
Asyntai pricing verified July 2026 in USD; EUR pricing is also available on our pricing page. For all other vendors, confirm current rates on their own pricing pages before any procurement paperwork is drafted.
The 6 options, compared
Asyntai
Hosted AI chatbot that answers using your own content — your Drupal pages plus any uploaded documents — added to Drupal with a single script through a module, a Twig template, or an asset library.
Asyntai takes the approach institutional web teams tend to prefer: the integration surface with Drupal is one asynchronous script tag, and everything else — knowledge base, AI configuration, conversation logs, team access — lives in a hosted dashboard. You point it at your Drupal site URL and it reads your public pages: program descriptions, service pages, policy documents, event listings, resource libraries. Documents that are not published on the site, such as internal handbooks or PDF reports, can be uploaded directly. When a visitor asks a question, the AI answers from that material and can link to the relevant page.
For a Drupal team, three properties stand out. First, the install does not enter your deployment pipeline: the Header and Footer Scripts module route is pure configuration, and the alternative is one line in html.html.twig. Second, there is no compatibility matrix to track — a core upgrade from Drupal 9 to 10, or 10 to whatever comes next, does not break a script tag. Third, language handling is automatic: the AI detects the visitor's language and replies in it, which matters enormously for international student recruitment and for public bodies with multilingual service obligations.
Plans scale by message volume rather than seats, which suits institutions where many staff may want dashboard access but chat volume is the real cost driver. Multi-site support maps well onto university structures: the Free plan covers 1 website, Starter 2, Standard 3, and Pro up to 20 — each with its own knowledge base and logs, relevant if you run a Drupal multisite for faculties or departments.
Pros
- Script-based install: no module deployment, no compatibility tracking across Drupal core upgrades
- Answers using your own content — Drupal pages plus uploaded PDFs and documents
- Automatic per-visitor language detection; widget UI localizable
- Free plan (100 messages/month, no card); paid plans are flat monthly fees by message volume
- Under 50KB, loads asynchronously; does not interfere with Drupal caching, Varnish or a CDN
- Escalation, human takeover, live monitoring and AI notifications on Standard+ for staffed service desks
- Multisite-friendly: up to 20 sites on Pro, each independently configured
Cons
- Configuration and conversation logs live outside Drupal, in the Asyntai dashboard
- Not a Drupal-native module — teams that require everything managed inside Drupal will prefer the contrib route
- Hosted service, so it must pass your institution's vendor review like any other SaaS
The open-source Drupal AI module route
Not a product but a toolkit: contributed modules on Drupal.org that connect Drupal to language-model providers, from which a development team assembles its own chatbot.
Drupal's contributed module ecosystem includes an active family of AI modules: framework modules that abstract over multiple language-model providers, plus companion modules for chatbots, search, content assistance and automation. A team installs them with Composer, supplies API keys for its chosen provider, indexes its content, and configures the chatbot behavior — all inside Drupal, version-controlled with the rest of the site.
The honest trade-off is that "free software" here means free as in licensing, not free as in effort. Someone on staff must assemble the pieces, tune retrieval quality, watch the API bill, test after every module update, and re-verify compatibility at each Drupal core upgrade. For institutions with a standing Drupal development team and a policy preference for open source and self-hosted data flows, that cost can be worth paying: you get complete control over where content is sent and how the assistant behaves. For teams whose developers are already fully booked, it becomes an unowned system within a year.
Pros
- Everything lives inside Drupal: config export, code review, version control
- Full control over data flows and choice of AI provider
- No vendor subscription — costs are API usage and staff time
- Aligned with open-source procurement policies
Cons
- Requires real Drupal development capacity to build and to keep running
- Module releases are version-specific; core upgrades add regression-testing work
- Retrieval quality, guardrails and multilingual behavior are yours to engineer
- Total cost of ownership is easy to underestimate
Zendesk
Enterprise customer-service platform whose messaging widget and AI agents can be placed on a Drupal site via a script; the chatbot is one component of a much larger suite.
Some large institutions already run Zendesk as their service desk — IT support, admissions inquiries, citizen requests. In that situation, adding its messaging widget to the Drupal site keeps chat inside the existing ticketing, routing and reporting infrastructure, which the support operation may value more than any property of the widget itself.
How it differs from Asyntai: Zendesk is a service-desk platform priced per agent, with AI capabilities layered onto the suite; check the vendor's current pricing. Asyntai is an AI chat layer priced by message volume, independent of how many staff have access. An institution that only needs the website to answer questions does not need to procure a full helpdesk to get it.
Pros
- Integrates chat with an existing Zendesk ticketing operation
- Suite-level reporting, routing and SLA tooling
Cons
- Per-agent pricing scales with team size, not chat volume
- Heavy procurement footprint if all you need is website Q&A
- No Drupal-specific integration; it is a generic web widget
Tidio
Hosted live-chat platform with an AI agent (Lyro), installed on Drupal via a JavaScript snippet; oriented toward small and mid-sized commercial websites.
Tidio bundles live chat, automation flows and its Lyro AI agent in one hosted tool. On Drupal it installs the same way any script does. Its product design and marketing center on ecommerce and SMB use — sales prompts, cart recovery, conversion nudges — which reads differently on a ministry or faculty website, but the underlying chat mechanics work anywhere a script tag does.
How it differs from Asyntai: Tidio's paid structure combines per-seat elements with separately metered AI conversations; check the vendor's current pricing. Asyntai prices by message volume with the AI included, and its automatic language detection is a first-class feature rather than an add-on consideration — a distinction that matters most on institutional sites serving many language groups.
Pros
- Free tier available for evaluation
- Live chat and AI agent in a single tool
Cons
- AI conversations metered separately from the base plan
- Product framing is commerce-oriented rather than institutional
Elfsight
A marketplace of embeddable website widgets — among them an AI chatbot — added to Drupal with a copy-paste embed code and managed from the Elfsight panel.
Elfsight sells a catalog of drop-in widgets (reviews, forms, galleries, chat and more), and its AI chatbot widget follows the same pattern: configure in the Elfsight dashboard, paste the embed code into your Drupal site. For teams that already maintain other Elfsight widgets, adding a chatbot from the same panel is low-friction. It is a widget product first, so the chatbot's depth — knowledge management, conversation operations, escalation workflows — is scoped accordingly.
How it differs from Asyntai: Elfsight is a general widget vendor with a chatbot in the catalog; check the vendor's current pricing, which is typically per widget. Asyntai is a dedicated AI chatbot service, with the knowledge base, retrieval from your own content, multilingual replies and support operations features as the core product rather than one item in a catalog.
Pros
- Very simple embed-code installation
- Convenient if you already manage other Elfsight widgets
Cons
- Chatbot is one widget among dozens, not the vendor's core focus
- Per-widget subscriptions accumulate across a large site portfolio
LiveChatAI
Hosted AI chatbot that builds its answers from content you import — website pages and documents — and installs on Drupal with a JavaScript snippet.
LiveChatAI occupies the same general category as Asyntai: a hosted AI assistant fed by your own site content and documents, with human-takeover capability, installed by script. On Drupal it has no platform-specific integration; you add the snippet as you would any third-party script. Its plans meter AI usage in tiers; check the vendor's current pricing.
How it differs from Asyntai: the categories overlap, so the comparison comes down to specifics an institutional buyer should test side by side: how each handles multilingual visitors (Asyntai detects and replies per visitor automatically), message allowances per dollar at your projected volume, multisite support (Asyntai Pro covers up to 20 websites under one account), and the operational features your service desk needs — escalation, monitoring, notifications.
Pros
- Same content-fed AI approach, so evaluation is directly comparable
- Script install works on any Drupal version
Cons
- No Drupal-specific documentation or integration path
- Usage-metered tiers require careful volume forecasting
How to choose: the institutional checklist
Strip away the feature grids and an institutional Drupal evaluation reduces to six questions. Put them in your RFP or your internal memo; they separate the options quickly.
- Module or script? A contributed module lives in your codebase: it goes through code review, dev–stage–prod deployment, and must be re-verified at every core upgrade. A script-based service (Asyntai, and most hosted vendors) is configuration: it can be enabled by a site builder, survives core upgrades untouched, and can be removed in one edit. Decide which model your governance actually supports before comparing anything else.
- Who reviews the data handling? Any hosted chatbot means visitor messages leave your infrastructure. Your data-protection officer will ask where conversations are processed and stored, how long they are retained, and how staff access is controlled. Ask every vendor the same written questions — and if you must keep all data flows in-house, the open-source module route is the only option on this list that fully satisfies that constraint.
- Multilingual by default or by configuration? A university answering prospective students from forty countries cannot pre-build forty language flows. Prefer a chatbot that detects the visitor's language and answers in it automatically, and check whether the widget chrome (buttons, placeholders, notices) localizes too.
- What is the year-two maintenance burden? The person evaluating chatbots today may not be employed there when Drupal 12 lands. Score each option on what survives staff turnover: hosted script services need essentially nothing; self-built module stacks need a named owner, a test plan and a budget line.
- Accessibility and front-end discipline. Public-sector and university sites carry accessibility obligations that a third-party widget becomes part of. Ask vendors for their accessibility documentation, test the widget with keyboard navigation and a screen reader on a staging site, and confirm the script loads asynchronously so it cannot degrade page performance budgets.
- Seats or messages? Institutions have many stakeholders who want to see the dashboard and a finite volume of actual chat. Per-agent pricing (Zendesk-style) charges you for the stakeholders; volume pricing (Asyntai-style) charges you for the conversations. Model both against your real numbers.
How to install an AI chatbot on Drupal
These are the real steps for Asyntai, drawn from our Drupal installation guide. Total time is a few minutes, and no code deployment is required for the recommended route.
Step 1 — Get your embed code. Create an account, then open your Asyntai dashboard and scroll to the Embed Code section. The code is a single script tag containing your personal widget ID.
Step 2 — Add it via the Header and Footer Scripts module (recommended):
- Log in to your Drupal admin panel
- Go to Extend and choose Install new module, then install and enable the Header and Footer Scripts module
- Go to Configuration → Development → Header and Footer Scripts
- Open the Footer tab and paste your Asyntai embed code into the footer scripts area
- Click Save configuration
Placing the code in the footer means it loads after your page content — the right behavior for a chat widget.
Alternative — edit the theme template: open themes/your_theme/html.html.twig, paste the embed code just before the closing </body> tag, save, and clear caches via Configuration → Performance → Clear all caches. Be aware that theme file edits can be overwritten by theme updates, so a child theme or the module route above is the more durable choice.
Alternative for developers — Drupal's asset library system: declare the widget script as an external library in your theme's *.libraries.yml (with async and the data-asyntai-id attribute), attach the library in *.info.yml, and clear caches. This keeps the integration inside your theme's declared assets, which some teams prefer for auditability.
Step 3 — Verify. Visit the site in an incognito window and confirm the chat button appears bottom-right and opens correctly. If it does not show, clear Drupal's cache first — with Drupal's aggressive caching, that resolves most cases.
What institutional Drupal sites use a chatbot for
Student, applicant and citizen questions
The questions arriving at a university or government site are repetitive in the aggregate and urgent to each individual: application deadlines, required documents, tuition and fee structures, permit procedures, office locations, eligibility rules. The answers already exist somewhere in your Drupal content — often several menu levels deep. A chatbot that answers using your own content collapses that navigation problem into one question and one direct answer, with a link to the authoritative page.
Multilingual service without multilingual staffing
International applicants and multilingual resident populations write in their own languages. Staffing a service desk across a dozen languages is unrealistic for most institutions; a chatbot that detects the visitor's language and responds in it provides a level of first-line coverage that no staffing plan matches, while your team handles escalations in the languages it actually speaks.
After-hours and peak-season coverage
Admissions traffic spikes at deadlines; citizen inquiries spike when a policy changes; and much of it arrives outside office hours or from other time zones. The chatbot holds the line at 2 a.m. and on the weekend before a deadline, and on Asyntai's Standard plan and above, conversations that genuinely need a human can trigger escalation and notification workflows so staff pick them up in the morning with context intact.
Deep-content wayfinding
Mature Drupal sites accumulate thousands of nodes across departments, programs and archives. Search helps users who know the right keywords; a conversational assistant helps everyone else — the visitor who asks "can I park on campus with a visitor pass?" without knowing which of your 4,000 pages holds the parking policy.
What you'll actually pay
Budget lines an institutional buyer can defend, verified for Asyntai and stated in principle for the rest:
- Asyntai: Free plan with 100 messages per month and no card required — enough for a genuine pilot on one site. Paid plans are flat monthly fees by volume: Starter $39/month (2,500 messages, 2 websites), Standard $139/month (15,000 messages, 3 websites), Pro $449/month (50,000 messages, up to 20 websites). Annual billing reduces each tier, and EUR pricing is available — details on the pricing page. No per-seat fees, so dashboard access for additional staff does not change the invoice.
- Drupal AI module route: no license fee, but budget for language-model API usage at your traffic level plus developer time for the initial build and ongoing maintenance — realistically the largest line item on this list for most teams.
- Zendesk: per-agent subscription tiers with AI capabilities in the suite; check the vendor's current pricing and model it against your agent count.
- Tidio, Elfsight, LiveChatAI: each publishes its own tiers — typically a free or entry tier plus paid plans metered by seats, widgets or AI usage. Check each vendor's current pricing at your projected volume before comparing.
Frequently asked questions
Does Drupal have AI?
Drupal itself does not ship with a built-in AI assistant. AI arrives through two routes: contributed modules from the open-source ecosystem that connect Drupal to language-model providers, or hosted chatbot services such as Asyntai that you add with a small script and manage outside Drupal core.
Is there a free AI chatbot for Drupal?
Yes. Asyntai's free plan includes 100 AI messages per month, covers one website, and installs on Drupal through the Header and Footer Scripts module or a theme template — no card required. Tidio and Elfsight also offer free tiers with their own limits; check each vendor's current terms.
How do I add an AI chatbot to a Drupal site?
For a hosted chatbot like Asyntai: copy the embed code from your dashboard, install the Header and Footer Scripts module, paste the code into the Footer tab under Configuration → Development → Header and Footer Scripts, and save. Alternatively, paste it into your theme's html.html.twig before the closing body tag and clear caches. Full steps are in the installation guide.
Which Drupal versions are supported?
Script-based chatbots such as Asyntai work with Drupal 7, 8, 9, 10 and newer, because they only require one JavaScript snippet in the page output. Contributed AI modules are version-specific: each release supports particular Drupal core versions, so confirm compatibility before committing to a module-based build.
Will an AI chatbot slow down a Drupal site?
A well-built widget should not. Asyntai's script is under 50KB and loads asynchronously after the page has rendered, so it does not block content. It also does not interfere with Drupal's page caching, Varnish, or a CDN — the layers institutional Drupal sites typically rely on for performance.
Can a Drupal chatbot answer in the visitor's language?
Yes, if the chatbot supports it. Asyntai detects each visitor's language automatically and replies in it, without requiring a separate configuration per language — useful for universities with international students and public bodies serving multilingual populations. The widget interface itself can also be localized.
Conclusion
The right chatbot for an institutional Drupal site is the one your team can still defend in front of the IT steering committee two years from now. Judged on that horizon — install path that fits governance, no version-coupling to Drupal core, multilingual answers from your existing content, volume-based pricing without seat penalties, and a free tier that permits a real pilot before procurement — Asyntai is our recommendation. Details and a live demo are on the AI chatbot for Drupal page.
The alternatives each have a legitimate lane: the open-source module ecosystem if you have Drupal developers and a mandate to keep every data flow in-house, Zendesk if the institution already lives in its service desk, and Tidio, Elfsight or LiveChatAI if their specific bundles match your situation — verify current pricing with each vendor. If you manage learning platforms alongside your main site, our comparison of AI chatbots for Moodle covers that adjacent decision, and for larger support operations there is a separate guide to AI chatbots for enterprise customer support.