MagicSchool AI

AI tool for teachers that helps create lessons, quizzes, and feedback.
It saves time and supports classroom planning.

Tool characteristics

MagicSchool AI is an education-focused platform designed to support both teaching and learning through a wide range of AI tools for teachers and students. Teachers can use it to create lesson plans, quizzes, rubrics, feedback, differentiated materials, and curriculum-aligned resources. They can also build personalised tool collections, access previous outputs through the Output History, and save chatbot conversations or uploaded materials in the Resource Library.

The platform also includes student-oriented tools and ready-to-use templates for activities such as language practice, tutoring support, reading aloud, writing assistance, and personalised feedback. Students can configure some tools according to their learning goals and use structured resources prepared by teachers. Some templates are limited to English, while multilingual and translation features can support more diverse classrooms.

MagicSchool AI also allows teachers to export content to Google Docs or Slides and integrate materials with learning-management systems such as Canvas. Overall, it supports more individualised, engaging, and privacy-aware classroom instruction, while helping teachers save time in planning, assessment, differentiation, and student suppo

MagicSchool AI can support learners in developing several language skills, including vocabulary, reading, listening, speaking, and writing. Its strongest support is in English, while other languages may be less consistently covered. The platform can generate word lists, simplify texts, create writing prompts, provide model texts, and offer personalised feedback adapted to learners’ proficiency levels.

MagicSchool builds on large language models (LLMs) from other providers using models from OpenAI and Anthropic (AI company behind the LLM Claude).

It incorporates fine-tuned prompting and domain-specific overlays. MagicSchool has tailored the generic LLMs for teacher workflow, adding special safety and ethical guardrails, formatting, templates, and user-flows to make the system teacher-friendly.

It is not clear which languages are supported on the platform. MagicSchool AI claims to support multilingual features. The platform includes a Text Translator tool, allowing users to translate materials into other languages. It also supports multilingual content generation. Users can generate materials in multiple languages for diverse classroom settings. The platform provides pedagogical support for creating a language inclusive environment in the multilingual classroom.

Most real-time possibilities only apply for English. Learners can receive feedback and instant suggestions on their writing, vocabulary use or reading. The platform allows spontaneous modification of materials (text difficulty, vocabulary lists) to suit learner needs. Learners can engage in interactive tasks and work with dynamic tools (choice boards, interactive prompts) generated live, helping them practise spontaneously.

Teachers can generate lesson plans, materials, worksheets, or assessments spontaneously based on the current class context, enabling rapid adaptation to different proficiency levels or learner groups during class.

Teachers and students can get quick translations of words, phrases, or sentences to support communication and comprehension during lessons. Through chat-based interactions, learners can practice writing and responding in the target language with instant feedback and corrections. By using the vocabulary tools learners can receive vocabulary lists, flashcards, and contextual examples immediately during lessons or study sessions.

Teachers can customise generated content (lesson-plans, worksheets, reading passages) to fit specific learner levels, language proficiency, and subject area. The platform allows differentiation and personalized learning paths to focus on areas needing improvement. Teachers can provide multiple versions of a text or quiz for learners at different language levels (beginner/intermediate/advanced) or with different needs.

Teachers can translate or adapt materials for multilingual classrooms (e.g., translate rubrics or texts into students’ native languages) and modify formats to support language learners.

The system offers adaptable templates and tools where teachers supply parameters (subject, objectives, language level, learner needs) and then tweak the AI-output to align with their pedagogy and classroom context.

Tasks and texts can be adapted in real-time to match their current proficiency, prior knowledge, and preferred language mode (e.g., simpler vocabulary, more scaffolding) which supports accessibility and personalised learning.

MagicSchool AI includes a range of assessment tools that help teachers create quizzes, tests, rubrics, and feedback activities more efficiently. Its Quiz Generator supports the quick preparation of assessment tasks based on topics, texts, or learning objectives, while the Rubric Generator helps create clear grading criteria for different types of student work.

The platform also supports formative assessment through tools such as Writing Feedback, which provides instant comments on learners’ written production, highlighting both strengths and areas for improvement. Assessment activities can be customised according to grade level, educational standards, or language level, making them useful for differentiated and multilingual classrooms.

For language learning, MagicSchool AI can support vocabulary practice, flashcards with images, audio and text, interactive quizzes, and speaking tasks with feedback on pronunciation and fluency. Overall, it helps teachers save time, diversify assessment formats, and monitor learners’ progress more regularly.

According to the accessibility statement, MagicSchool AI is committed to ensuring that the website is accessible for all individuals by following the Web Content Accessibility Guidelines (WCAG) 2.0. MagicSchool AI states that their ongoing efforts include regular audits and updates to improve accessibility features, ensuring that all content is navigable and understandable. MagicSchool provides supports such as read-aloud, text-to-speech, web accessibility adjustments, and language translation features applicable for ELL (English Language Learner) / MLL (Multilingual learner) populations as well as students with special needs.

Teachers and learners can access the platform free, and the interface is designed to be user-friendly for beginners. Users can work with simple prompts and get guidance when using the tools. Integration and export options (Google Docs, Microsoft Word, LMS connectors, SSO) give teachers freedom to embed the platform into existing workflows and devices. Teachers can quickly adjust tasks, change language levels, export materials to familiar formats, and support multilingual or differentiated classrooms.

The platform emphasises data privacy and safe usage. It claims robust security, customized data-privacy agreements (DPAs), and guardrails against misuse (e.g., bias, personal data exposure). It collects personal data when user create an account: First and last name, school or organisation name, role or job title, email address, and location information including state/province and country. It also collects data about user’s device and software, domain name, and time stamp of visit. It also collects analytic data by using third-party analytic tools (e.g., Google Analytics), and cookies.

According MagicSchool AI, no AI provider has the right to train models on any data processed on MagicSchool’s behalf. Any data sent through APIs (Application program interfaces) is retained for up to 30 days solely for abuse and misuse monitoring before being permanently deleted—unless required by law to retain it longer.

Target Group

Features

It supports vocabulary building, reading, writing, listening, and speaking practice. It also helps learners receive feedback, simplify texts, and practise language through interactive activities.

For teachers, it helps develop skills in lesson planning, assessment design, feedback writing, differentiation, and classroom communication. It also supports the creation of quizzes, rubrics, and adapted materials.

Its support is more limited. It can help with translation, text simplification, multilingual communication, and drafting content, but it is not a specialised professional tool for advanced interpreting or high-precision translation.

MagicSchool AI supports engagement by providing personalised tasks, instant feedback, and accessible learning materials that make participation easier and more motivating. Its interactive activities can help students stay involved, practise more actively, and feel more confident in the learning process.

MagicSchool AI improves engagement by helping them create more interactive, relevant, and differentiated activities, which can increase student participation and attention. It also reduces preparation time, allowing teachers to focus more on classroom interaction and responsive teaching.

MagicSchool AI offers only limited engagement, mainly through basic translation and multilingual support tools.

Ease of use comes from guided tools built for classroom use, such as tutoring, summarising, rewriting, translation, writing feedback, and step-by-step support. These features are designed to help students interact with AI in a structured and accessible way, without needing advanced technical skills.

MagicSchool AI is generally easy for teachers to use because it offers a large set of ready-made tools for lesson planning, quizzes, rubrics, feedback, text adaptation, and many other classroom tasks in one platform. Its design is clearly teacher-oriented, so educators can start from templates and examples instead of building prompts from scratch.

MagicSchool AI is easy to use for basic translation and multilingual support, but it is not designed for specialised professional language mediation tasks.

MagicSchool AI offers a more structured and safer environment than general-purpose AI tools, which improves reliability in classroom use. Its accuracy can still vary, though, especially for factual, nuanced, or recent content, so learner-facing outputs should be monitored by teachers and not treated as automatically correct.

MagicSchool AI can be considered moderately reliable and reasonably accurate as a teaching support tool because it is designed for school workflows and includes guardrails, moderation, and educator oversight. However, MagicSchool itself states that teachers must review outputs for accuracy, bias, and classroom fit, and that AI responses may sometimes be incorrect or not fully up to date.

MagicSchool AI is not designed for high-precision professional language mediation.

AI explainability is more limited but still present in a guided form. MagicSchool emphasizes AI literacy, monitored student use, and structured classroom support, which helps students understand that the system is a learning tool with safeguards rather than an independent authority. However, learners are not given deep technical explanations of how the model reasons, so explainability is mainly pedagogical rather than algorithmic.

MagicSchool AI offers moderate explainability for teachers because it makes its educational purpose, safeguards, and human-review model fairly clear. The platform explains that outputs are shaped by educator-focused guidance, examples, and the information the user provides, and it also states that teachers remain responsible for reviewing outputs rather than treating them as self-explaining or automatically authoritative.

MagicSchool AI does not explain professional linguistic or interpreting decisions in depth.

Autonomy is also limited and guided. MagicSchool’s student-facing tools are designed to support learning within teacher-defined boundaries, with moderation, visibility, and classroom controls that keep student use aligned with instructional goals rather than allowing fully independent AI-driven activity.

MagicSchool AI offers limited autonomy for teachers because it works mainly as a support system rather than an independent decision-maker. It can generate lessons, assessments, feedback, and differentiated materials, but the platform is explicitly designed around educator judgment, human oversight, and teacher control, which means teachers remain responsible for reviewing, adapting, and approving outputs.

MagicSchool AI has low autonomy, as it supports basic multilingual tasks but is not designed for independent professional language mediation.