Free AI grader for teachers

20 graded items, once. The allowance belongs to the account for as long as the account exists. It does not reset on the first of the month and it does not expire, so there is no reason to rush through it or to save it up. Everything the account has ever graded counts against it. When the 20 are used, grading more means upgrading.

The same engine as the paid plans. Free is a volume limit, not a weaker model. A page graded on the free tier is marked exactly as it would be on Pro.

Worksheets and essays on one allowance. There is no separate essay quota. A math page and an essay draw on the same 20.

One photographed or scanned page. A five-page worksheet PDF is five items. A two-page handwritten test is two.

An essay bills by length. It costs the larger of its page count and one item per 500 words of typed text. A 1,200-word essay pasted as text is three items; a two-page handwritten essay is two, because the page count is the floor. 500 words is what a single-spaced typed page holds, so one item is one page either way.

Twenty short worksheets are 20 items. Twenty 1,500-word research essays are sixty, so a free account grades about six of those. Pick what you test it on accordingly.

• Worksheets, problem sets and exit tickets: photographed off paper or uploaded as a PDF.
• Quizzes and tests: multiple choice, short answer and worked problems on one page.
• Handwritten math: each problem worked out, the wrong step named, equivalent forms accepted.
• Science, English, world languages and social studies: any page of finished work, no template.
• Essays: scored criterion by criterion on one of 86 state and national rubrics, or your own.

The constraint is the format, not the subject. It reads a page of finished student work photographed with a phone; there is no template to print, no bubble sheet and no scanner. For the math case in detail, see AI math grader for handwritten work. For essays, the free AI essay grader page covers the rubrics and compares every free essay-grading tier in the category.

• Marks on the student’s own page, next to each answer, not a score in a separate window.
• A reason beside every mistake, in plain English a student can act on. On math, that is the step that went wrong, with a short worked solution under the mark.
• Equivalent forms accepted. Each problem is worked out rather than matched to a key, so ½, 0.5 and 50% all count, with or without units.
• Unreadable answers flagged, never guessed. A faint minus sign or an illegible digit comes back to you with the grader’s best reading, to confirm or correct in one tap.
• Your override on everything. Change any mark, edit any comment, adjust any credit, and nothing reaches a student until you release it. It grades; the grade stays yours.

Volume, and the classroom scaffolding around it. Pro is $30 a month for 400 graded items a month and adds class rosters, your own saved rubric library, per-student and per-class analytics, gradebook and CSV export, and Google Classroom and Canvas sync. Max is $50 a month for 1,000, sized for a full secondary teaching load. The grading itself is identical on every tier. See the plans.

It does not renew. 20 items is the whole free allowance for the account, ever. If you need a class set a week, that is a paid plan.

It does not replace you. It marks a first pass and explains what it marked; every score is editable and the grade is yours before it is released.

It does not score figures. A freehand diagram, graph or geometry construction is read far less reliably than working and text. The working beside a figure is marked; the figure itself you check yourself.

It is not a plagiarism or AI-writing detector. We do not do similarity checking and nothing here substitutes for a tool that does.

It will not read every page. Faint pencil, heavy crossing-out and photos taken at an angle all cost accuracy, which is why unreadable answers are flagged rather than marked wrong.

There is no published accuracy benchmark for math. The measured figures we publish are for essays: On 78 real student essays that Texas released with official examiner scores, nine in ten came back within one point of the examiner and quadratic weighted kappa was 0.78, against 0.70 as the commonly accepted bar for an automated marking system. The page says that is essays only, rather than implying more.

We hold no SOC 2 certification. Institution plans are FERPA and COPPA aligned with a custom data processing agreement; if your district requires a certification we do not have, that decides it, and we would rather say so here than at the end of a procurement call.

• Free AI essay grader for teachers: what each free essay-grading tier in the category includes, dated.
• AI math grader for handwritten work: the wrong step named, equivalent forms accepted.
• AI grading for handwritten work: every subject, photographed off paper.
• How to grade a whole class set of worksheets: end to end, and what thirty pages cost in items.
• What is the best AI grading tool for teachers?: six situations, including the three where the answer is not us.
• Every AI grading tool compared: head to head, with each vendor’s real price.
• Writing rubric library and printable worksheets: free, no account.