Preply Arabic Tutoring vs Duolingo Chinese (Mandarin)

Preply has no Arabic curriculum of its own — all content comes from individual tutors, who range from structured certified instructors to informal conversation partners. The platform covers the full spectrum of Arabic varieties (MSA, Egyptian, Levantine, Gulf dialects, Quranic Arabic, Tajweed), which is a genuine strength for learners with specific dialect goals. Progress tracking tools, AI Lesson Insights, and between-lesson Q&A support add some structure, but the absence of a dedicated Arabic vocabulary section on the mobile app (unlike English or French) and no community or group features leave meaningful content gaps. The ceiling is high with the right tutor; the floor is whatever the lowest-rated tutor brings.

The Arabic tutor pool on Preply is deep — over 6,600 tutors drawn from Egypt, Lebanon, Jordan, Morocco, Saudi Arabia, and elsewhere, covering both native and near-native speakers across all proficiency levels. Reviewer analysis of tutor profiles consistently praises patience, personalised lesson planning, and cultural context. The platform categorises tutors into Super Tutors (highest-rated), Professional Tutors (certified), and general tutors, giving learners a rough quality signal. Preply does not enforce formal teaching credentials across all tutors, however, so quality variance is real and the burden of vetting falls on the learner through trial lessons and careful profile screening.

Arabic lessons start at $2/hour and average around $10/hour across the full pool, with native-speaker tutors from Egypt averaging $16/hr, Lebanese $20/hr, and Jordanian $17/hr — competitive with italki for the same quality tier. The trial lesson is discounted and covered by a 100% money-back guarantee if unsatisfied, lowering initial risk. Value is meaningfully dented by the mandatory package structure (lessons must be bought in batches, not one at a time), the subscription auto-renewal, and the fact that Preply retains 18–33% commission from tutors, pushing some quality instructors toward platforms that pay more fairly. For committed weekly learners the per-lesson math is strong; for occasional learners the credit model creates real friction.

Scheduling, messaging, and tutor-matching interfaces are reported as smooth and intuitive across independent reviews. The weak spot is billing and post-cancellation support: multiple verified complaints on PissedConsumer and Trustpilot describe auto-renewal charges appearing after cancellation, unused lesson credits expiring without refund, and customer service that escalates to AI chatbots rather than responsive human agents. Preply holds a 4.3/5 overall Trustpilot score from 21,500+ reviews, but 1.5/5 on the more complaints-focused PissedConsumer (90% unfavourable). The gap reflects a support experience that works well for standard cases and fails badly when something goes wrong with money.

The core use case — regular 1-on-1 conversation with a native Arabic speaker who adapts to your dialect goal and corrects your output in real time — is the clearest path from passive vocabulary knowledge to actual spoken Arabic, and multiple independent sources confirm this. Learners who commit to two or more sessions per week consistently describe faster progress than app-only routines. The dialect coverage (MSA for formal/academic goals, Egyptian for broad intelligibility, Levantine for conversational use, Gulf Arabic for the region) maps precisely onto the real-world contexts Arabic learners typically target. A 2025 Preply study found learners who completed 24+ lessons over 12 weeks progressed three times faster than typical learning timelines.

The course introduces pinyin and pairs hanzi with sound reasonably well in the early lessons, and vocabulary exposure is broad. But Mandarin exposes Duolingo's thin content faster than European languages: reviewers repeatedly describe near-absent tone training, no character writing or stroke order, and sentences that are sometimes unnatural. The Chinese tree was also locked in mid-2022, so known errors and broken audio were frozen rather than fixed.

There is no instructor. The method is implicit pattern-matching, and for Mandarin that breaks down badly — grammar is almost never explained, and the four tones (the single most important feature for being understood) are effectively ignored. Reviewers consistently say the app expects you to absorb rules and pronunciation you were never actually taught. For a language this distant from English, the hands-off approach is the core teaching weakness.

The core course is genuinely free, which is its strongest argument — zero cost exposure to pinyin, characters, and basic vocabulary. Super at roughly $7-13/month only removes ads and adds hearts; reviewers agree it does not fix the structural gaps in tones, grammar, or speaking. The value lives entirely in the free tier, and even there several reviewers conclude the time is better spent on Mandarin-specific apps.

Gamification remains Duolingo's standout strength even for Mandarin. Streaks, points, and reminders genuinely build a daily habit, and reviewers with 400-day streaks credit the app with getting them to practise every day. The catch is the well-documented ceiling: recognition keeps improving while real ability — especially tones and conversation — stalls, and the streak can become the goal in place of the learning.

Duolingo support is email-only and slow, and for Chinese specifically the situation is worse: the course was locked in mid-2022, which froze the community discussion threads, the user corrections, and the broken or missing audio. The third-party Mandarin-blog ecosystem partially fills the gap, but the official Chinese course is effectively in maintenance silence rather than actively supported.

This is the weakest area. Tones are barely trained, speaking exercises only ask you to repeat scripted sentences, and several reviewers report completing the entire tree and still being unable to hold a basic Mandarin conversation or produce tones accurately. Because tones carry meaning, weak tone training directly limits real-world usability. It is a vocabulary and habit supplement, not a path to spoken Mandarin on its own.