Smarter IB revision, powered by AI
AI graders for your TOK essay, Extended Essay and Internal Assessment - built by subject specialists and machine-learning PhDs, trained on real IB-awarded scripts, and marked to the official rubric.
Get Started Free- 5,000+ students
- 3 AI graders
- 480+ resources
- 100+ guides
- First grade free, no card
Instant feedback on every IB essay
Each grader is purpose-built for its subject and tuned to that subject's own official rubric and examiner reports - covering TOK, all 10 Extended Essay subjects and 27 Internal Assessment subject-levels. You get a predicted grade and a criterion-by-criterion breakdown in minutes, not weeks.
- TOK Essay Grader (/10) - Every prescribed title. Your 1,600-word essay, marked against the four TOK assessment instructions.
- Extended Essay Grader (/34) - All 10 subjects. Your 4,000-word research essay, marked against IB criteria A to E.
- Internal Assessment Grader (/14-40) - 27 subject-levels (HL & SL). Your subject coursework, marked against the official per-subject rubric.
Frequently asked questions
What is MyRevisionAgent?
An AI-powered IB study platform built by ML PhDs and IB graduates who scored 40+. We offer three AI graders (TOK, Extended Essay and Internal Assessment), 480+ premium study resources and 100+ in-depth guides covering every part of the IB Diploma.
Which IB essays and coursework can I grade?
All of them. Our TOK Essay Grader scores your essay /10, the Extended Essay Grader covers 10 subjects and marks /34, and the Internal Assessment Grader supports 27 subject-levels (HL/SL) with per-subject rubrics from /14 up to /40.
What makes your graders different from ChatGPT or Claude?
Frontier LLMs are powerful but they weren't built to mark an IB essay. Our graders use our own custom-built models combined with classical ML (gradient-boosted trees, ridge regression), per-subject feature weighting, and an independent statistical validator that catches outliers. They've been tested against thousands of examiner-marked TOK, EE and IA submissions.
How accurate is the AI grading?
Predictions are calibrated against IB-awarded marks and ship with a typical-error range and confidence score, not a single guess. On a held-out EE test set the predicted grade lands within one band of the examiner's 96% of the time, with a mean absolute error of about 3.6 marks out of 34. When our grading model and the statistical validator disagree by more than two points, we surface why.
Who built MyRevisionAgent?
A team of PhD-level machine-learning researchers and quantitative engineers, working with experienced IB examiners and graduates who scored 40+. We bring the same rigour to grading IB essays that quant teams apply to financial models.
Is it free?
Yes - all 480+ study resources are free after creating an account, every one of the 100+ guides is free to read, and we add one free grading token to your account on signup (usable on any grader: TOK, EE or IA). Additional grades are paid via tokens, with packs from £4.99.
Are my submissions used to train your models?
No. Your essays are processed for your grade and stored only to your account history. They're never used for model training or shared with third parties.
What subjects and IB elements do you cover?
The full IB Core - TOK, the Extended Essay (10 subjects) and Internal Assessment (27 subject-levels across HL and SL). Our 480+ study resources and 100+ guides also cover IB subject content, exams, UCAS, Oxbridge interviews and study tools.
Who is MyRevisionAgent for?
IB Diploma students aiming for 40+, and teachers who want curated, exam-aligned resources and AI-assisted feedback to scale 1-to-1 marking.