Teaching · 6 min read

By Mohamad Mohamad · 19 August 2026

Free reading material for language teachers

Handing a whole class the same book is the part of extensive reading that costs money. Thirty licences, or thirty paperbacks, or a subscription with a per-seat price and a login for every student. The material itself is often free — the friction is licensing, level, and the account each platform wants your students to create.

This is a working guide to the free sources, what each one is genuinely good for, and the one thing none of them solve.

Where the free material actually is

Everything below is public domain or openly licensed, which means you may copy it, hand it to thirty students, put it on a school server, and edit it, without asking anyone.

• Project Gutenberg — the largest collection, and the one most likely to have a specific title. Files are plain and occasionally rough: inconsistent quote marks, transcriber's notes left in the text, the odd scanning error. Fine for reading, worth a skim before you assign it.
• Standard Ebooks — volunteers take Gutenberg texts and typeset them properly: corrected typography, real chapter structure, modernised spelling where the original was merely archaic. A much smaller catalogue, noticeably better files. This is where to look first if the title exists.
• Wikisource — strong on documents, speeches, short pieces and translations that never got a commercial edition. Useful when you want one text rather than one book.
• LibriVox — public-domain audiobooks, volunteer-read, quality varies by reader. Pairing a chapter of audio with the same chapter of text is one of the few genuinely free ways to build listening and reading together.
• Your public library — not public domain, but usually free to your students and carries contemporary books the sources above cannot. Worth checking what a library card already entitles a class to before buying anything.

The catch nobody mentions: public domain means old

A book enters the public domain decades after its author dies. That is the whole catalogue's defining constraint. The free sources are rich in nineteenth and early twentieth century prose and almost empty of the register your students actually hear.

This matters more than it sounds. Older prose is not merely quaint — it is measurably harder. Longer sentences, a wider vocabulary, and words that have since fallen out of use but still look like ordinary words, so a student does not know to be suspicious of them.

Here is the shape of it in the 170 English books on this site, all of which are public domain, and all of which were measured the same way:

The right-hand column is the one to read carefully. Research on extensive reading generally puts comfortable independent reading somewhere around 95% to 98% known words — the level at which a reader can infer the rest and keep going. Not one of the 170 books reaches 95% for a B1 reader. The most accessible manages 92.4%; the hardest sits at 78.4%, which is roughly one unknown word in every five. Even measured against a B2 vocabulary, the best book on the shelf only reaches 94.4%.

So the honest position is this: a B1 student handed a free classic is not reading comfortably. They are decoding. That is not an argument against public-domain books — they are free, they are legal, and many of them are wonderful. It is an argument for knowing the number before you assign the book, and for giving students something to do about the gap.

How to check a book before you assign it

Three questions, in order of how much time they take.

Is it long enough to be worth starting, and short enough to be finished?

Total word count is the honest measure, not page count, which depends on the edition. A class reading twenty minutes a week gets through roughly 40,000 words a term at a modest pace. A 200,000-word novel is a year's project you are unlikely to finish.

What proportion of the words will they already know?

This is the number in the table above, and it is the one that predicts whether students give up. If you have no coverage figure for a book, a rough substitute is to take a page from the middle — not the first page, which authors polish — and count the words you would expect the class not to know. More than five in a hundred is hard going.

Is the subject something they will keep going for?

Interest compensates for a surprising amount of difficulty, and nothing compensates for boredom. A student who wants to know what happens next will tolerate a much lower coverage figure than a student reading because it was assigned.

The measurement method used on this site is written up in full, including what it does not capture, if you want to apply the same approach to books of your own.

Using ClickBook with a class, free

ClickBook is a free reading app for iPhone, iPad and Android, so it works in a class where students have different phones. It is built for the gap the table above describes: a reader who is one unknown word in ten away from comfortable, and needs those words answered without leaving the page.

What is relevant if you are the teacher rather than the reader:

• No accounts. Students do not sign up, do not give an email address, and are not tracked. There is nothing for you to administer and no student data to be responsible for.
• It works offline. The model that explains words runs on the phone itself. No school wifi, no data plan, nothing sent to a server. This also means it works on the bus, which is where a good deal of reading homework actually happens.
• 408 measured books are free to download — public domain, across English, French, German, Spanish, Italian, Portuguese and Dutch, listed on this site as EPUB files that open in the app. Three classics ship with the app itself; the rest are one download away. The 170 English titles carry an estimated CEFR level, so you can point a class at a level rather than a guess.
• Explanations come in over 40 languages. In a class where students do not share a first language, each one can read the same English book and get explanations in their own.
• Students can load their own EPUB files, so anything you find at the sources above can go straight in.

What it does not do

Worth knowing before you build a lesson around it.

• No teacher dashboard. No class lists, no assignment tracking, no reports on who read what. The privacy position and the reporting features are the same decision seen from two sides — the app cannot tell you what a student read because it does not know either.
• No built-in exercises. It is a reader, not a coursebook. Comprehension questions, vocabulary tasks and follow-up are yours.
• The explanations are generated. A model running on a phone is small, and it is occasionally wrong. For a reader keeping momentum through a chapter this is an acceptable trade; for anything a student will be assessed on, it is not a reference work.

A starting sequence that works

If you want to try one cycle without rebuilding your scheme of work: pick a short book, well below the level you think the class can handle — the goal of the first one is finishing, not stretching. Check the coverage figure. Give it three weeks with twenty minutes of class time a week and no comprehension test at the end, because the first book is establishing that finishing a book in another language is possible at all. Ask only what happened and whether they liked it.

Then let them choose the second one. That is the point at which extensive reading either takes hold or does not, and choice is most of what decides it.