How LitBook creates and reviews reading lessons.
LitBook uses AI to create reading passages matched to a learner's language, CEFR level, topic and length. Every generated lesson is checked for structure, level fit, language quality and answerability before it reaches the learner; lessons that still fail after one repair attempt are blocked and refunded.
Last reviewed 23 September 2026.
How does LitBook create reading passages?
- Define the learning target. The request specifies language, CEFR level, topic, length and the learner's instruction language.
- Generate a private draft. The passage, answer choices, hints and answer key are created away from the learner-facing lesson.
- Run deterministic checks. Code checks the schema, word range, question mix, answer options, leakage and whether every answer is grounded in the passage.
- Review language and pedagogy. An independent reviewer checks grammar, naturalness, CEFR fit, clarity and whether each question has one defensible answer.
- Repair once or fail closed. A failed lesson gets one repair attempt and every check runs again. A second failure is blocked and refunded rather than shown.
The workflow is described in more technical detail in our AI reading-passage comparison and model benchmark.
What do LitBook's CEFR levels mean?
CEFR is the Common European Framework of Reference for Languages. LitBook uses its progression—from A0/A1 beginner reading to C2 proficient reading—to control vocabulary, sentence structure, abstraction and question style. The level is a practical difficulty setting, not an official certification. Read our plain-language CEFR guide or consult the Council of Europe's official CEFR resources ↗.
How are guides and comparisons published?
Answer first
Every guide gives a direct answer before introducing LitBook, then provides the evidence and trade-offs needed to act on it.
Primary sources first
Research claims use original papers, official documentation or first-party product pages whenever possible.
Useful before promotional
Calls to action follow the answer and lead to the closest relevant level test or practice page rather than interrupting it.
Comparisons disclose our interest
LitBook is named as our product, limitations are stated, and competitors are recommended when they fit better.
Testing claims require evidence
We describe exactly what was checked and never imply hands-on testing that did not happen.
Updates must be material
An Updated date changes only after facts, recommendations, structure or original evidence materially change.
Common questions
- How does LitBook create reading passages?
- LitBook uses generative AI to draft a passage for the learner's chosen language, CEFR level, topic and length. The draft is not shown immediately: code checks its structure, length, answers and passage grounding, then an independent reviewer checks language quality and level fit.
- What CEFR levels does LitBook support?
- LitBook supports A0 through C2 in the app. A0 is for first-contact reading, A1–A2 for basic users, B1–B2 for independent users and C1–C2 for proficient users. The labels guide difficulty; they are not a replacement for an official CEFR exam.
- How does LitBook check AI-generated lesson accuracy?
- Each lesson must pass deterministic checks for structure, word range, answer options and passage grounding, plus an independent language-quality review. LitBook allows one repair attempt; if the lesson still fails, it is blocked and the learner's credits are refunded.
- How are LitBook product comparisons researched?
- Comparisons use current first-party product pages for features, language support and pricing. LitBook discloses that it is our product, states its limitations and recommends a competitor when that competitor better fits the reader's job.
- How does LitBook handle corrections and updates?
- Material errors are corrected promptly. A visible updated date changes only when facts, recommendations, structure or original evidence change—not for cosmetic edits. Current product details are rechecked against first-party sources on the displayed review date.
Corrections and review
Product features and prices are checked against first-party pages on the displayed update date. If you spot an error, contact us through the support page; material corrections are made promptly and reflected in the page's updated date.
Contact supportUse the methodology to compare tools
Apply the same level-control, grounding, feedback and privacy criteria in our AI reading-comprehension tools comparison or review the narrower LingQ alternatives guide.