The best live chat software for a language school

Where language schools actually lose enrollments

Start with the leak, not the tool. The conversion step is the placement test or trial class, not the chat itself. Chat earns the placement booking, the placement earns the enrollment into a forming group. Every failure on this page is a failure at one of those two handoffs.

The visitor writes in their own language, after hours

A school website is usually written in one language, often the language taught, or English. The visitor is not. They compare schools from another country, in another timezone, and they ask the question that decides everything: what level am I, what does it cost, and when does the next group start. If the widget answers in the wrong language, or gives a course list instead of a level path, the visitor moves on. That is the whole leak in one line, and it is why reply language matters more here than in any other vertical we cover.

Widget language is not reply language

Most tools translate the widget interface: buttons, labels, and greetings. That is not the same as answering in the visitor language, from the school own pages, with levels, prices, and start dates checked in the languages students actually use. We score those two things separately on this page, and we say plainly where a vendor page did not let us verify reply language. A school that buys on widget translations alone will find out, from real visitors, that the answers still arrive in English.

A group is not a seat until it forms

A language group does not run until enough students share a level and a schedule. That single fact shapes the whole page. A bot that confirms a group, promises a seat, or fixes a start date is inventing an enrollment the school has not made. The honest chat answer is the forming rule itself: groups start when enough students at the same level share a schedule, start dates move, and the next step is a placement test. Schools that hide that rule get angry emails later. Schools that state it up front get visitors who arrive already understanding the wait.

We found no clean owner quote for language schools at build time, so this page carries none. No tool was trialed or tested for this page beyond reading its public pages.

What a language school actually needs from chat

Eight requirements, drawn from how a language school front desk actually works: inbound from abroad, placement testing, group formation, and three different buyers. Every tool below is scored against these eight, and nothing else.

1. Multilingual inbound, not just a multilingual widget

The visitor writes in their own language and the AI must answer in that language, from the school own pages, with key answers (levels, prices, start dates) checked in the languages students actually use. Score reply language and widget-interface language separately. They are not the same thing, and a tool that only translates buttons fails this requirement even if its pricing page leads with a language count.

2. Placement-test handoff, not placement testing

Chat collects goal, current level as self-assessed, target language, and preferred format (group vs private, in-person vs online), then hands a complete lead plus a placement-test booking to staff. Chat may run a rough self-assessment. The official level is set by the school test and a human, never by the bot. A tool that cannot hand off cleanly to the placement-test booking page is a worse pick even if its chat is good.

3. The group-formation boundary

Chat explains that groups start when enough students share a level and schedule, and that start dates move. It must never confirm a specific group will run, has a seat, or starts on a date unless connected to live enrollment data. Link to the placement-test and enrollment pages instead. This rule belongs in the tool instructions, not just in staff training, because after-hours visitors meet the bot first.

4. Group vs private routing

The same enquiry splits on budget, pace, and schedule. Chat must surface the real trade-off (group price and fixed schedule vs private price and flexibility) from published prices, then route the undecided visitor to an adviser, not push one format. A bot that always pushes the group course is selling, not advising, and it will misroute the visitor who needed private lessons to hit a deadline.

5. International, visa, and documents boundary

Chat explains which documents the school asks for and links to the official source. It must never advise on visa eligibility or promise a visa outcome. Document chase and payment steps stay in the school own enrollment process. This is a hard line, not a tone choice: a wrong visa answer in chat can cost a student a term and a school its reputation.

6. Adult, parent, and employer buyers, with a safeguarding hard line

The same school serves adult learners, parents of minors, and employers buying for staff. Chat must route each buyer type differently: employer and minor-safeguarding enquiries to a named person, adult self-enrollment to placement booking. Safeguarding hard line for minors: never collect a child full name, date of birth, or health information in chat. Any tool setup that asks for those fields in a pre-chat form fails this requirement outright.

7. Small-team cost, in each tool own AI unit

The typical school is a director plus a handful of teachers, often part-time. Compare flat or workspace pricing vs per-seat pricing honestly, and compare AI budgets in their own units, because they are not identical: Tidio counts only human replies as billable; Crisp budgets AI credits; LiveChat counts resolutions; Intercom charges $0.99 per Fin outcome; Freshchat counts sessions. Note multilingual volume too: every language multiplies the knowledge base to maintain, so a cheap tool with a knowledge base nobody maintains is not cheap for long.

8. Multi-location and multi-language variants

Separate schedules and levels per site and per language matter. A school teaching Spanish in one city and French in another, or running the same language across two sites, needs per-location and per-language knowledge that does not bleed across. Note which tools handle per-location and per-language knowledge cleanly, and say where a vendor page did not let us verify that separation.

Comparison: all eight tools against language-school requirements

Prices are annual billing unless labeled monthly, re-checked November 4, 2026, except where the method section notes an earlier check. Channel and translation coverage is from vendor pages read November 4, 2026.

How to read the AI allowances. LiveChat counts a resolution as one solved question, and multiple questions in one chat still count as one. Tidio counts a billable conversation only when a human agent replies, so Lyro AI replies do not consume the quota, but Lyro conversations are metered separately in their own quota. Freshchat counts a session as all AI interactions within 24 hours of the first message. Crisp counts automated conversations against an AI credit budget. Zoho counts bot chat sessions per month, with operator sessions counted separately. These are different units, so compare them as budgets, not as identical counts.

The pick: Zoho SalesIQ Professional

Zoho wins on the requirement that decides a language-school purchase: multilingual inbound at a price a small school can actually pay. Basic at $7 per operator a month billed annually already includes chat translation and messaging channels including WhatsApp and Instagram, plus a chatbot builder with 25,000 bot sessions a month. Professional at $12.75 per operator a month billed annually adds auto-translated bot content and resources, 5 chatbots, 50,000 bot sessions a month, and unlimited operator chat sessions. For a school teaching two or three languages, that last point is the pick: the bot content itself can be auto-translated, instead of the school hand-writing every placement question in every language before launch.

The placement shape fits too. A placement handoff is a structured flow, not a free-form essay: goal, self-assessed level, target language, format, then a booking link. A chatbot builder handles that flow well, and 25,000 to 50,000 bot sessions a month is far above what a small school will use, so volume is not the constraint. Seats are the constraint, and even there Zoho stays cheap: three operators on Professional cost $38.25 a month billed annually, and three on Basic cost $21 a month. The free plan covers 3 operators with 100 operator chat sessions a month, enough to test whether placement enquiries arrive in chat before paying. Per-website business hours and departments help a school with more than one site keep schedules apart, though per-location knowledge separation was not verified from the pages checked, so test that separation before promising it.

Two honest limits. First, Zoho AI on the pages checked is a chatbot builder and, on Enterprise, an Answer Bot, not a headline generative agent that writes free-form answers from school pages. A school that wants that style of AI should look at the runner-up, not pretend Zoho is it. Second, dollar prices on the vendor page did not render in our November 4 read, so the dollar figures here carry the October 30 check and are flagged that way in the method section. Neither limit changes the pick at small-school size, because translation plus a structured placement flow, at $38.25 a month for three operators, beats a stronger AI that answers in the wrong language or costs three times as much.

Runner-up: Tidio

Tidio is the pick for a school that wants the AI to write answers from its own pages, not follow a chatbot flow it built by hand. Starter at $24.17 a month billed annually covers a workspace, not a seat, with up to 10 seats on self-serve plans, so a director plus part-time teachers pay the workspace price, not a seat price each. Add Lyro from $32.50 a month billed annually for 50 Lyro conversations, and the combined cost is about $56.67 a month billed annually. The free plan covers 50 conversations and a one-off 50 Lyro conversations, enough to test whether placement enquiries are real for a specific school before paying. The 7-day trial needs no card.

Two honest limits decide why Tidio is second here, not first. Reply-language coverage by language was not verified from the pages checked, so a school teaching three languages must test Lyro answers in each language, with levels and start dates checked, before launch. And the entry Lyro quota is only 50 conversations a month, so a school with heavy inbound in intake season will need a larger Lyro quota, whose stepped prices were not re-read from the vendor page today and are not quoted here. For a single-language school with light volume, Tidio may still be the better fit. For the typical multi-language small school, Zoho translation on Professional is the safer default.

When a language school should pick LiveChat instead

LiveChat is a mature, dependable human-plus-AI tool, and its Growth plan is genuinely good for a larger school: 200 AI resolutions a month, unlimited conversation history, and, as a vendor claim, multilingual support with a choice of 48 languages and dedicated agent groups per language. That last feature is the boundary that matters here: a school running several languages across several sites, with a dedicated admissions person per language, can assign agent groups per language instead of routing everything through one inbox. For a typical single-site small school it loses on price and on the Essential gap: multilingual support sits on Growth, not Essential, so the realistic multilingual entry is $79 per user a month billed annually. A two-person team on Growth pays $158 a month billed annually, before any WhatsApp marketplace app fee, against $38.25 a month for three operators on Zoho Professional. At larger size the $158 is a smaller share of revenue and per-language agent groups matter more than the lowest price. That is the boundary: small school, Zoho; larger or multi-location school, LiveChat. The vendor page also notes that educational institutions can ask about additional discounts, which is a question to ask, not a promised discount.

Visit LiveChat

The LiveChat button above is an affiliate link. Every other tool link on this page is a plain homepage link.

The other five, briefly

Crisp: flat price, strong translation listing, higher entry

Crisp lists LiveTranslate in more than 50 languages, unlimited conversations on every plan, and flat workspace pricing, so a viral intake post cannot spike the bill. Essentials at $95 a month covers 10 seats and adds the omnichannel inbox with WhatsApp Business and Instagram DMs, with about 450 automated conversations from its $25 AI credit budget. For a school whose inbound is heavy and unpredictable, that flat model is attractive. For a typical small school it costs more than twice the Zoho Professional team price, and the plan scope for its translation features was not verified from the page extract checked, so confirm that scope before buying on translation alone.

Freshchat: strong AI, but multilingual sits on Pro

Freshchat free plan covers up to 10 agents, and Growth at $19 a month billed annually adds social messengers including Instagram plus text messaging channels. Freddy AI Agent costs $49 per 100 sessions after a one-time 500 free sessions on paid plans, which is a workable AI budget for a school testing automation. The catch for language schools is tier placement: multilingual conversations are listed on Pro at $49 a month billed annually per agent, not on Growth. A two-agent school buying for translation therefore starts at $98 a month billed annually before any Freddy sessions, which is hard to justify against Zoho Professional at $38.25 for three operators.

Intercom: capable, priced for larger teams

Intercom Essential is shown at $19 a seat a month against a $29 list price on the page checked, and Fin AI Agent costs $0.99 per outcome on top. WhatsApp is charged per message sent. A multilingual Help Center sits on Advanced, and AI auto-translation conversations appear in the Copilot add-on. A multi-location school group with dedicated support staff may fit here. A single small school will pay more than it needs to, and will pay for translation through an add-on rather than getting it in the base plan.

Zendesk: only if you already live there

Zendesk $19 Support Team plan includes no live chat, so a school buying for chat starts at Suite Team, $55 a month billed annually per agent, or $69 monthly. Suite Professional is $115 a month billed annually, $149 monthly. Its vendor page lists support for 80+ languages for its AI agents, but how that maps to a small school placement flow was not verified from the pages checked. A school already running Zendesk for support has a reason to stay. A school buying chat alone does not.

Olark: unlimited AI at an enterprise price, language coverage unverified

Olark Standard AI Website Agent at $400 a month billed annually includes unlimited conversations with no token costs and no overages, plus one human seat, with extra seats at $29 a month. That flat unlimited model is genuinely attractive for a high-volume school, and lead capture and qualification are listed features. At $4,800 a year it is priced for a group or a very busy school, not a single small school, and multilingual reply coverage was not verified from the pages checked, so a language school cannot buy it on language grounds from this page.

The math: what placement-season volume actually costs

Take a small school fielding about 50 AI-handled placement enquiries and 60 human-reply conversations a month in intake season. This is an example, not a statistic. Watch what happens to each tool price at that volume, in each tool own unit.

Zoho SalesIQ: Three operators on Professional pay $38.25 a month billed annually. Bot sessions are 50,000 a month and operator sessions are unlimited on Professional, so 50 AI-handled enquiries and 60 human conversations sit far inside the allowances. Volume does not move the price here. Seats do.

Tidio: Starter at $24.17 a month billed annually covers 100 human-reply conversations, which covers 60. Lyro from $32.50 a month billed annually covers 50 Lyro conversations, which covers 50. Together that is about $56.67 a month billed annually. If AI enquiries rise above 50, the Lyro quota must step up, and stepped prices above 50 were not re-read from the vendor page today, so budget that step before intake season, not during it.

Crisp: Essentials at $95 a month includes about 450 automated conversations from its AI credit budget, which covers 50 with room to grow, and unlimited human conversations cover 60. Mini at $45 includes about 90 automated conversations, which also covers 50, but Mini lacks the omnichannel inbox, so the realistic Crisp entry for a school answering Instagram and WhatsApp in one inbox is Essentials, not Mini.

LiveChat: Growth at $79 per user a month billed annually includes 200 AI resolutions, which covers 50. One user pays $79 a month billed annually, two users pay $158. Essential at $19 a month billed annually includes only 10 resolutions, so it does not cover 50 without extra 50-packs at $49.50 each, and it lacks the multilingual support that Growth includes. The lesson is general: on LiveChat, buy the plan whose allowance and language features cover the volume, do not bolt packs onto the small plan.

Freshchat, Intercom, Zendesk, Olark at the same volume: Freshchat Pro for translation costs $98 a month billed annually for two agents before Freddy sessions, and Freddy adds $49 per 100 sessions after the one-time 500 free sessions. Intercom at the shown $19 a seat plus $0.99 per Fin outcome comes to $38 for two seats plus $49.50 for 50 outcomes, before WhatsApp per-message charges, but translation sits in an add-on. Zendesk Suite Team for two agents costs $110 a month billed annually. Olark costs $400 a month billed annually regardless of volume. The spread at this modest volume, from $38.25 to $400, is the whole argument for pricing at team size and language need, not at headline plan names.

Chat vs the suite: where one ends and the other begins

Many language schools already pay for a platform that runs calendars, billing, placement-test records, or enrollment funnels. Buying a standalone chat tool on top of it can mean paying twice for the same inbox, or worse, running two placement flows that disagree. Check the platform first.

If you run a school platform or an admissions funnel

Studio and school platforms of the Mindbody and TutorBird style run calendars, billing, and scheduling. Admissions funnel products of the LinguaEnroll style run placement-test booking, group-class formation, and document chase. Chat is the conversation layer in front of them, not a replacement for any of that. The funnel product holds the group-formation rule and the enrollment record. Chat holds the first conversation and the handoff to it.

The handoff test

Judge any chat tool on one test: a visitor asks, in their own language, what level they are, what a group course costs, and when the next group starts, and the tool must collect goal, self-assessed level, target language, and format, explain the forming rule, and hand a complete lead plus a placement-test booking to staff, without confirming a group, a seat, or a visa outcome. A tool that passes the chat test but drops the lead between chat and the booking page fails the school, even if its widget looks good.

Do not score chat tools on placement testing itself

The test belongs to the school. Chat may run a rough self-assessment, and it must say so in those words, so the visitor does not mistake a few chat questions for an official level. The official level comes from the school test and a human. A vendor that sells chat as a placement test is selling the wrong product for this job, however good its chat is.

Setup checklist: the first week with chat on a language school site

This is the outcome work. A chat widget with an empty knowledge base is a prettier contact form. Load it with the school real answers, in the languages students actually use, before launch.

Load levels, not just course names

Levels taught, what each level means in plain words, group vs private prices, in-person vs online formats, and the placement-test booking link. The AI must answer from these pages and say it does not know when the answer is not there. A guessed level or price is worse than no answer, because the visitor enrolls on the wrong course.

Write the group-start rule into the tool, word for word

Groups start when enough students share a level and schedule, and start dates move. The bot must never confirm a specific group will run, has a seat, or starts on a date unless connected to live enrollment data. Test this rule with example messages in each language before launch, not after a real visitor is promised a group that never forms.

Write the visa and documents boundary into the tool

Which documents the school asks for, and a link to the official source for visa questions. No advice on visa eligibility, no promised outcome. Document chase and payment stay in the school own enrollment process, not in chat.

Write the safeguarding handoff rule before any automation

Parent enquiries and employer enquiries go to a named person. Adult self-enrollment goes to placement booking. Never collect a child full name, date of birth, or health information in chat, and remove any pre-chat form field that asks for them.

Check answers per language, not once in English

Test the same five questions in every language the school recruits in: level path, group price, private price, next group start rule, and placement-test booking. If an answer is wrong or arrives in English in any language, fix the knowledge base for that language before launch. Every language multiplies the knowledge base to maintain, so budget that maintenance, not just the launch.

Set the after-hours path explicitly

After hours, the widget should answer routine questions in the visitor language, capture the lead with goal, level, language, and format, and point to the placement-test booking page. It should not promise a callback time the school cannot keep, and it should not promise a group start it cannot guarantee.

Measure three numbers for a term

Placement-test bookings that started in chat, placement tests completed, and enrollments into forming groups. If the first number is high and the second is flat, the handoff is broken, not the widget. If the second is high and the third is flat, the group-formation wait is the problem, and the fix is clearer forming-rule messaging, not more chat.

Frequently asked questions

How can you check this page for yourself?