Preply Scraper

Extract Preply tutor profiles and student review text into two clean datasets — hourly rate, rating, lessons taught, languages, education, certificates and full review text — using Preply's own GraphQL endpoint, with no browser and no residential proxy.

• Reviews as a dataset, not a buried string. Most Preply tools expose profile fields only. This one writes one row per review into a separate reviews dataset, joinable to tutors on tutorId.
• No browser, no residential proxy. Talks to https://preply.com/graphql/v2/ directly, so there is no Cloudflare browser challenge to pay for or wait out.
• Cheapest verified Preply tutor dataset on the Apify Store at $1.90 per 1,000 tutors on the FREE tier, against $3.00 per 1,000 for the closest verified alternative.

Main Use Cases

• Preply tutor database — build a structured catalogue of tutors by subject, language and price band
• Hourly tutor rate benchmarking — compare hourlyPrice and trialLessonPrice across hundreds of tutors to map the market
• Tutor review analysis — collect full review text with scores, dates and tutor replies for sentiment and quality research
• Online tutor marketplace research — study supply, pricing and quality to plan a competing marketplace
• Language tutor lead lists — filter by certified tutors, native speakers, country and language to build a qualified prospect pool
• Tutor credential verification — extract educations, certificates and isCertified / isSuperTutor flags

How It Works

1. In search mode, the Actor pages Preply's tutor directory for a subject slug and language, following startPage up to maxPages.
2. Each listing is parsed into a flat TutorRecord with detailLevel: "listing".
3. For each tutor the Actor optionally calls the tutor(id) profile query, which fills in the richer fields — educations, certificates, workExperiences, reviewDistribution, responseTimeHours, videoIntroUrl — producing detailLevel: "profile".
4. When includeReviews is on, each review is parsed into a ReviewRecord and pushed to the reviews dataset, carrying tutorId as the join key.

Requests go straight to Preply's GraphQL endpoint over plain HTTPS.

Extract Preply Tutor Data Without a Browser or Residential Proxy

Preply sits behind Cloudflare, which is why most scrapers for it need a headless browser and a residential proxy — and why the cheapest options still cost the most per result. This Actor queries Preply's own GraphQL API instead. No browser, no proxy, no challenge to solve, and a lower per-result price than the browser-based alternatives.

Reviews are deliberately kept in a separate dataset. A tutor with 400 reviews still produces exactly one tutor row plus 400 review rows, so no single row ever carries an unbounded nested array.

Input

Field	Type	Required	Default	Notes
mode	string	no	search	search walks the tutor directory. profiles fetches only the URLs/IDs you supply.
subjectSlug	string	no	english	Subject/category slug, e.g. english, mathematics, programming. Used in search mode.
languageCode	string	no	en	Site language prefix.
startPage	integer	no	1	First directory page to read. Raise this to resume deeper pagination.
maxPages	integer	no	2	Maximum directory pages to walk. Default is deliberately small — raise it for real collection.
sort	string	no	rating	One of rating, reviews, popularity, price_lowest, price_highest.
priceRange	string	no	—	Price range filter, passed through to the directory query.
tutorCountry	string	no	—	Restrict tutors to a country of birth.
searchText	string	no	—	Free-text keyword filter on the directory listing.
certifiedOnly	boolean	no	false	Only professional/certified tutors.
nativeSpeakersOnly	boolean	no	false	Only native speakers of the selected language.
alsoSpeaks	string	no	—	Require an additional spoken language.
startUrls	array	no	—	Direct tutor profile URLs. Used in profiles mode.
tutorIds	array	no	—	Direct tutor IDs. Used in profiles mode.
includeReviews	boolean	no	false	Also fetch reviews into the reviews dataset. Costs extra requests per tutor.
maxItems	integer	no	20	Maximum tutors to collect.
proxy	object	no	—	Proxy configuration. Not needed for the default GraphQL path.

Output

Two datasets are written.

Default dataset — tutors (one row per tutor):

tutorId, name, subjects, subjectSlugs, hourlyPrice, currency, trialLessonPrice, rating, reviewsCount, lessonsTaught, languages (name, code, level), bio, headline, availability (hasAvailability, activeStudents, lessonsBookedLast48h), profileUrl, country, countryCode, avatarUrl, lastSeen, isOnline, yearsOfExperience, preplyTenureYears, responseTimeHours, isCertified, isSuperTutor, verifiedTutor, videoIntroUrl, tags, metadataTags, educations, certificates, workExperiences, reviewDistribution, sourceUrl, detailLevel, scrapedAt

detailLevel tells you how complete a row is: "listing" (directory-level fields only) or "profile" (the full profile query ran).

reviews dataset (one row per review):

reviewId, tutorId, tutorName, tutorProfileUrl, score, text, createdAt, language, isEdited, verifiedByCs, relevanceScore, reviewerUserId, reviewerLessons, reviewerLastLessonAt, reviewerLearningGoal, tutorReplies (text, createdAt, language), sourceUrl, scrapedAt

Join the two datasets on tutorId.

Example Input

{
  "mode": "search",
  "subjectSlug": "english",
  "languageCode": "en",
  "sort": "rating",
  "certifiedOnly": true,
  "startPage": 1,
  "maxPages": 5,
  "maxItems": 100,
  "includeReviews": true
}

Example Output

{
  "tutorId": "184023",
  "name": "Amelia Hart",
  "subjects": ["English - Business English", "English - Conversation"],
  "subjectSlugs": ["business-english", "conversation"],
  "hourlyPrice": 27,
  "currency": "USD",
  "trialLessonPrice": 14,
  "rating": 4.98,
  "reviewsCount": 412,
  "lessonsTaught": 3180,
  "languages": [
    { "name": "English", "code": "en", "level": "Native" }
  ],
  "bio": "Cambridge-certified English teacher with 9 years of experience teaching professionals.",
  "headline": "Business English that gets results",
  "availability": {
    "hasAvailability": true,
    "activeStudents": 6,
    "lessonsBookedLast48h": 2
  },
  "profileUrl": "https://preply.com/en/teachers/amelia-hart",
  "country": "United Kingdom",
  "countryCode": "GB",
  "isOnline": true,
  "yearsOfExperience": 9,
  "preplyTenureYears": 5,
  "responseTimeHours": 2,
  "isCertified": true,
  "isSuperTutor": true,
  "verifiedTutor": true,
  "educations": [
    { "degree": "MA English Literature", "school": "University of Cambridge", "dates": "2012-2015" }
  ],
  "certificates": [
    { "name": "CELTA", "dates": "2015" }
  ],
  "workExperiences": [
    { "position": "Corporate Trainer", "company": "Deloitte", "dates": "2017-2021" }
  ],
  "sourceUrl": "https://preply.com/en/teachers/amelia-hart",
  "detailLevel": "profile",
  "scrapedAt": "2026-02-11T09:14:22.301Z"
}

And one row from the reviews dataset for the same tutor:

{
  "reviewId": "99281744",
  "tutorId": "184023",
  "tutorName": "Amelia Hart",
  "tutorProfileUrl": "https://preply.com/en/teachers/amelia-hart",
  "score": 5,
  "text": "Amelia prepared me for a client presentation in two weeks. Her feedback on my intonation was genuinely useful.",
  "createdAt": "2026-01-28T10:03:00Z",
  "language": "en",
  "isEdited": false,
  "verifiedByCs": true,
  "reviewerLessons": 24,
  "reviewerLearningGoal": "Business English",
  "tutorReplies": [
    {
      "text": "Glad it went well — good luck with the presentation!",
      "createdAt": "2026-01-28T12:41:00Z",
      "language": "en"
    }
  ],
  "sourceUrl": "https://preply.com/en/teachers/amelia-hart",
  "scrapedAt": "2026-02-11T09:14:22.301Z"
}

Pricing

Pay-per-result. A tutor costs $1.90 per 1,000 on the FREE tier. Setting includeReviews: true also bills for each review at $1.00 per 1,000. Fetching a full profile costs more request budget than reading the directory listing alone, so set maxItems deliberately when you need profile-level fields.

Integrations & Automation

• Apify API — run it from Python, Node or any HTTP client, and read the datasets directly.
• Schedules — refresh lastSeen, activeStudents and ratings daily; use a weekly schedule for market snapshots.
• Webhooks — trigger a downstream job when a run finishes.
• MCP — callable by AI agents through the Apify MCP server.

Related Actors

• ahmed_jasarevic/link-preview-api — pull the Open Graph and JSON-LD metadata for tutor profile URLs as a cheap supplementary signal
• ahmed_jasarevic/otta-scraper — comparable job-listing extraction with salary and equity fields, if your research spans tutor and tech hiring markets

FAQ

Why use this actor instead of Preply's own API?

Preply does publish an OpenAPI document at https://preply.com/openapi.yaml, but it describes the scoped plugin/partner integration surface, not the tutor directory. The directory and profile data used here come from Preply's internal GraphQL endpoint at https://preply.com/graphql/v2/.

Do I need a residential proxy?

No. This Actor calls the GraphQL endpoint directly, so the default configuration needs no proxy at all. A proxy field exists if you want to add one.

Do I get review text, or just review counts?

Full review text. Reviews are written to a separate reviews dataset — one row per review, including score, text, createdAt, reviewerLessons, reviewerLearningGoal and any tutorReplies.

What is the difference between listing and profile rows?

detailLevel is "listing" when the tutor came only from a directory page, and "profile" when the full profile query also ran. Profile rows add educations, certificates, workExperiences, reviewDistribution, responseTimeHours, isOnline, preplyTenureYears and videoIntroUrl. Listing rows return null for those fields rather than inventing values.

Which mode should I use — search or profiles?

search walks the directory for a subjectSlug, which is how you build a broad database. profiles fetches only the exact startUrls or tutorIds you pass, which is how you enrich a list you already have.

What are alternatives to this actor?

The main alternatives are browser-and-proxy-based Preply scrapers. They tend to cost more per result, need a residential proxy, and expose review text only if you pay extra. One further option is a narrow email-only extractor, which does not give you profile data at all.

Why is my priceRange filter ignored?

Preply's directory query does not apply the price filter server-side in all locales, so the Actor filters on price locally. Expect slightly fewer results than an unfiltered page.

For AI Agents & LLM Apps

Purpose: returns Preply tutor profiles as tabular rows in the default dataset and student reviews as rows in a named reviews dataset, joinable on tutorId.

Minimal working input — smallest input that produces a valid run:

{ "subjectSlug": "english", "maxItems": 5 }

Search variant — paginate the directory and sort by price:

{
  "subjectSlug": "english",
  "languageCode": "en",
  "sort": "price_lowest",
  "startPage": 1,
  "maxPages": 5,
  "maxItems": 200
}

Profiles variant — fetch specific tutors only, with reviews:

{
  "mode": "profiles",
  "tutorIds": ["184023"],
  "includeReviews": true,
  "maxItems": 5
}

Output fields — tutors: tutorId, name, subjects, subjectSlugs, hourlyPrice, currency, trialLessonPrice, rating, reviewsCount, lessonsTaught, languages, bio, headline, availability, profileUrl, country, countryCode, isOnline, yearsOfExperience, preplyTenureYears, responseTimeHours, isCertified, isSuperTutor, verifiedTutor, educations, certificates, workExperiences, reviewDistribution, detailLevel, sourceUrl, scrapedAt. Reviews (named dataset reviews): reviewId, tutorId, tutorName, tutorProfileUrl, score, text, createdAt, language, reviewerLessons, reviewerLearningGoal, tutorReplies, sourceUrl, scrapedAt.

Behaviors an agent should know

• Reviews land in a separate named dataset, not in the tutor row. Read both and join on tutorId.
• mode defaults to search. In profiles mode, only startUrls and tutorIds are used — subjectSlug and the filters are ignored.
• maxPages defaults to 2, which yields very few tutors. Raise it alongside maxItems or you will under-collect.
• includeReviews adds a request per tutor and bills per review. Leave it off unless reviews are actually needed.
• Rows with detailLevel: "listing" have null in the profile-only fields. Treat null as "not fetched", not "absent".
• priceRange is filtered locally, so result counts can be lower than the directory page size.
• No proxy is required on the default path.

Billing: pay-per-result — $1.90 per 1,000 tutors on the FREE tier, plus $1.00 per 1,000 reviews when includeReviews is on.

Legal & Compliance

This Actor is an independent tool. It is not affiliated with, endorsed by, or sponsored by Preply.

It reads only publicly accessible tutor listings, public tutor profiles, and public student reviews from Preply's own GraphQL endpoint. It does not bypass authentication, does not attempt to defeat Cloudflare challenges, and does not use CAPTCHA solving.

You are responsible for your own compliance with Preply's Terms of Service and with applicable data-protection law. The output contains personal data — tutor names, biographies, education history, work history and student review text including reviewer names and learning goals. Under GDPR and comparable regimes you need a lawful basis for processing this data, and you must handle it accordingly.

Please do not use extracted personal data for unsolicited commercial outreach in violation of applicable law, including CAN-SPAM, GDPR or TCPA.

SEO Keywords

preply scraper, preply tutor data, preply tutor profile, preply tutor database, preply api, preply graphql api, preply reviews, preply tutor ratings, hourly tutor rate comparison, preply lesson price data, online tutor database, tutor market research, tutor review text dataset, language tutor leads, esl tutor price research, tutor certification verification, tutor subject catalog, preply alternatives, preply teacher data export, teach english online data, build a language tutor marketplace, preply tutor search