AI Tutor for A-Level Maths: 2026 UK Sixth-Form Parent Guide

Your Year 12 child came home last Friday with their AS mock results. The maths grade was a C. They want to study economics at a <a href="/blog/ai-tutor-a-level-2026-uk">Russell Group university</a>. UCAS tariff points for economics require a B at A-Level minimum, ideally an A. Their school teacher said "they need to work harder at integration." You have typed "A-Level maths tutor" into Google at 11pm. The first three results are private tutors at £40 to £50 an hour. Here is the honest 2026 guide on what an A-Level maths AI tutor can and cannot do, what it actually costs, and the five buying criteria that matter for a sixth-form maths student working toward their target grade.

What is an A-Level maths AI tutor?

It is software that prepares a sixth-form student for A-Level maths exams using artificial intelligence. The AI:

• Diagnoses what the student already knows across the full A-Level maths specification — <a href="/blog/ai-tutor-a-level-2026-uk">pure maths, statistics, and mechanics</a> — not just the GCSE content they finished last summer
• Identifies the specific subtopics where marks are being lost. For A-Level maths, this is often integration by parts, parametric equations, proof by contradiction, conditional probability, moments and couples, or work-energy problems with non-conservative forces
• Picks the next question based on those gaps, not a fixed scheme of work or a generic textbook chapter order
• Adapts difficulty up or down in real time as the student improves or struggles
• Remembers every wrong answer across every session, so it never wastes time on a topic the student has already mastered
• Tracks progress against the actual exam board specification — AQA, OCR, Edexcel or CIE — not a generic "A-Level maths" content list. The AQA Paper 1 tests different content weightings than OCR A Paper 1. The AI knows the difference.

It is not a chatbot. It is not a YouTube video playlist. It is a diagnostic engine with a question bank mapped to the specific A-Level maths specification the student is sitting, with mark scheme alignment for every question.

Why use AI instead of a private A-Level maths tutor?

A private A-Level maths tutor costs £40 to £70 per hour. The UK national average from the Tutor Association 2026 survey is £48 per hour for A-Level maths, higher than GCSE because the content is harder and qualified maths graduates are scarce outside London. A serious student sees a tutor once a week, for an hour, between September and June. That is roughly 36 hours of contact time across the academic year, costing £1,728. The tutor teaches from a generic textbook and a past-paper pack. They do not know what the student did in school that week. They cover the same topics in the same order regardless of need.

An AI tutor costs £8 to £15 per month. It works with the student every day, for as long as they want to sit and work. Across a 9-month academic year that is £72 to £135. It remembers every wrong answer. It focuses on what the student actually needs.

For A-Level maths specifically, the AI has a structural advantage over a weekly tutor. A-Level maths is cumulative — every topic in Paper 2 builds on Paper 1 content. A student who did not fully understand C2 integration techniques in Year 12 cannot do Year 13 differential equations. A weekly tutor does not have the bandwidth to track which foundational gaps are still open. The AI tracks every question the student has ever answered, marks it against the spec, and surfaces the gap when it matters. For a Year 13 student targeting an A or A* in AQA maths, the AI will drill proof by contradiction, vectors in 3D, parametric-differentiation links, and moments with non-perpendicular forces across the final term — the exact topics that differentiate the A from the B in the AQA Paper 2 mark distribution.

What does Grademy's A-Level maths AI tutor cover?

Pure mathematics — full coverage of AQA, OCR, OCR B (MEI), Edexcel, CIE:

• Proof: proof by deduction, proof by exhaustion, proof by contradiction, proof by induction
• Algebra and functions: partial fractions, functions and their graphs, composite and inverse functions, modulus functions, transformations
• Coordinate geometry: Cartesian and parametric equations, curves and their equations, circle geometry
• Sequences and series: binomial expansion, arithmetic and geometric series, sigma notation, recurrence relations
• Trigonometry: radians, trigonometric identities, trigonometric equations, double-angle formulas, harmonic form, small angle approximations, further trigonometric equations
• Exponentials and logarithms: exponential functions, logarithmic functions, laws of logs, natural log and e, exponential growth and decay modelling
• Differentiation: differentiation from first principles, product rule, quotient rule, chain rule, implicit differentiation, parametric differentiation, second derivatives
• Integration: standard integrals, integration by substitution, integration by parts, integration using partial fractions, integration of trigonometric functions, definite integrals, areas under and between curves, volumes of revolution
• Numerical methods: locating roots, iteration, Newton-Raphson, trapezium rule, Simpson's rule, numerical integration, numerical solution of differential equations
• Vectors: 3D vectors, vector equations of lines and planes, angles between lines and planes, distances, scalar product, vector product (AQA further maths only)

Statistics — full coverage of AQA, OCR, OCR B (MEI), Edexcel, CIE:

• Statistical sampling: simple random sampling, stratified sampling, systematic sampling, sampling distributions
• Data presentation and interpretation: measures of central tendency and spread, box plots, histograms, cumulative frequency, coding, outliers
• Probability: mutually exclusive and independent events, conditional probability, Venn diagrams, tree diagrams, probability distributions
• Statistical distributions: discrete uniform, binomial, geometric, Poisson, normal, continuous uniform
• Statistical hypothesis testing: null and alternative hypotheses, one-tail and two-tail tests, critical values, significance levels, Type I and Type II errors, confidence intervals
• Quantities and units, regression and correlation: scatter diagrams, correlation coefficients, regression lines, Spearman's rank, Kendall's tau, least squares, residuals

Mechanics — full coverage of AQA, OCR, OCR B (MEI), Edexcel, CIE:

• Kinematics in one and two dimensions: constant acceleration equations, projectiles, vector methods, motion with variable acceleration (AQA further maths)
• Forces and Newton's laws: free body diagrams, equilibrium, motion with constant and variable force, connected particles, pulleys, rough inclined planes
• Moments: moments of forces, couples, equilibrium of a rigid body, centre of mass, toppling
• Work, energy and power: work-energy principle, conservation of energy, kinetic energy, gravitational potential energy, work done against resistive forces, power
• Further kinematics and dynamics (AQA further maths, OCR B further): motion in a circle, simple harmonic motion, damped and forced oscillations, resonance

Mechanics or Statistics options — AQA offers Mechanics Major or Statistics Major as Year 2 options. OCR offers Mechanics a or Statistics a. The AI routes the student to the correct option based on the school's choice.

Past paper practice — full sets of past papers for AQA, OCR, OCR B (MEI), Edexcel A-Level maths from 2018 onward, including specimen papers and the new 2025 baseline papers. Marked against the official mark scheme, with feedback on every question and identification of the specific mark scheme point lost.

Required practicals (OCR B MEI and CIE) — statistical and mechanics practicals with mark scheme alignment. AQA and Edexcel A-Level maths has no required practicals, so this is non-applicable for those exam boards.

What does the AI actually do day to day?

Here is what a typical week looks like for a Year 12 A-Level maths student using Grademy's AI tutor in November of Year 12, eight months before the final exams. The student is on the AQA exam board, targeting a B grade, struggling with integration.

1. Monday, 30 minutes. Student opens the AI and starts a session. The AI pulls up a diagnostic quiz on integration by parts — last attempted three weeks ago, scored 4 out of 8. Today's session has 8 integration questions, starting with simple single-step integration by parts, progressing to repeated integration by parts for sin(x)e^x type questions, then a defiite integral requiring substitution followed by parts. Student gets 5 out of 8. AI records this and moves on.

2. Tuesday, 25 minutes. AI notices the student has not practised logarithms and exponentials since the GCSE bridge work in September. It drops 6 questions into the session: solve 3^x = 45, sketch y = log_2(x), state the domain of y = ln(2x+1). Student gets 4 out of 6. AI diagnoses that the student is not fluent with the natural log laws. It schedules 5 natural log law questions for Wednesday.

3. Wednesday, 20 minutes. Student does the natural log law drill. Gets 4 out of 5. AI marks this gap as "closing" and schedules a final check in 14 days (spaced repetition).

4. Thursday, 45 minutes. Student sits an AQA Paper 1 mini-paper from 2023. Score: 62 out of 100. AI marks the 38 wrong answers against the AQA mark scheme, identifies 12 as careless errors (silly mistakes, not reading the question properly, sign errors), 18 as genuine knowledge gaps (vectors in 3D, parametric equations, proof by contradiction), and 8 as partially correct (correct method, lost marks on presentation or final step). The 18 genuine gaps become the focus for the next two weeks.

5. Friday, 30 minutes. Student chooses to work on vectors because the school covered vectors the previous week and they want to consolidate. AI gives 10 vector questions. Score: 7 out of 10. AI notes the 3 wrong answers were all parametric equations questions — a sub-topic not yet taught at school. AI schedules parametric equations for the half-term week.

6. Saturday, 60 minutes. Student runs a focused mechanics session before Monday's school mechanics test. AI pulls up 12 questions on connected particles and rough inclined planes — the exact topics from this week's schoolwork. Score: 9 out of 12. The 3 wrong answers are all friction coefficient questions where the student picked the wrong value of g (using 9.81 instead of 10, or vice versa, depending on the AQA paper convention). AI flags this and notes it for next mechanics session.

That is 3 hours 30 minutes of focused, adaptive practice across the week. A private tutor in the same week would have seen the student once, for an hour, and covered 2 topics from a textbook.

When is an AI A-Level maths tutor the wrong choice?

AI tutors do not work well for every sixth-form maths student. Be honest about this.

• Oxbridge or medicine applicants. Oxford and Cambridge STEP papers, and the BMAT for medicine, test problem-solving under time pressure in ways that AI drilling alone does not prepare a student for. These students still need a specialist tutor who can walk through STEP I and STEP II problems. AI is excellent for the main A-Level syllabus but is a complement, not a replacement, at this level.
• Students with severe maths anxiety or executive function issues. A weekly 1:1 with a human tutor provides accountability and emotional regulation that AI cannot. A student who cannot sit down to open the AI on their own will not benefit from the best AI in the world.
• Students who need extended worked examples for understanding. AI gives feedback on right and wrong answers, but if a student does not understand a concept from the feedback, they need a human who can re-explain using a different approach. AI is excellent at drilling once a concept is understood; it is weaker at the first-explanation step for students who struggle with abstract mathematical thinking.
• Further Maths A-Level. Further Maths (AQA, OCR B MEI, Edexcel) covers content beyond the standard A-Level — complex numbers, matrices, polar coordinates, hyperbolic functions, Maclaurin series, differential equations at a deeper level. AI coverage of Further Maths is patchy and improving but not yet comprehensive. For Further Maths students, AI is a useful complement but not the primary tool.
• Coursework-heavy specifications. CIE A-Level maths has a paper 5 (probability and statistics) and paper 6 (modelling) that some schools use for coursework. AI supports these but the teacher's specific school conventions still matter.

5 buying criteria for an A-Level maths AI tutor

1. Exam board specificity. The AI must support AQA, OCR, OCR B (MEI), Edexcel, and CIE as distinct specifications. A generic "A-Level maths" content list is not enough. AQA Paper 1 weighs pure maths at 100% while OCR A Paper 1 splits pure maths and mechanics/statistics. The AI has to know which exam board your child's school uses and route accordingly.

2. Mark scheme alignment. The AI must mark every question against the actual exam board mark scheme, not a generic "correct/incorrect" judgement. In AQA A-Level maths, a question worth 6 marks requires the AI to identify which mark scheme point the student lost (M1 for method, A1 for accuracy, B1 for an independent follow-through). This granularity tells the student exactly what to revise.

3. Diagnostic granularity. The AI must identify subtopic gaps, not just chapter gaps. "Student is weak at integration" is not useful. "Student loses marks on repeated integration by parts involving e^x × trig functions, but is fluent on single-step integration by parts and integration by substitution" is useful. The diagnostic has to operate at the mark-scheme-point level.

4. Spaced repetition. A-Level maths topics compound. A student who mastered logarithms in September but has not practised them since will lose fluency by February. The AI must resurface previously-mastered topics at the right interval (typically 14, 30, and 60 days) to maintain retention without wasting time on already-mastered content.

5. Exam technique feedback. Beyond knowing the maths, A-Level maths students lose marks on presentation, "show that" questions, and final-answer form (exact values, surds, 3 significant figures, etc). The AI must mark against the exam board's specific conventions for these. AQA penalises missing units in mechanics answers. OCR penalises unsimplified surds. CIE penalises decimal approximations where exact forms are required. The AI must know these conventions and mark accordingly.

Cost comparison — UK 2026

Option	Cost	Frequency	Diagnostic depth	Mark scheme alignment
Private A-Level maths tutor (UK avg)	£48/hr	Weekly 1hr session	Topic level	Generic
Private A-Level maths tutor (London)	£55-£75/hr	Weekly 1hr session	Topic level	Generic
Online tutor marketplace (MyTutor, Tutorful)	£30-£45/hr	Weekly 1hr session	Topic level	Generic
Grademy A-Level maths AI	£12/mo	Daily, on demand	Subtopic + mark-scheme-point level	Full alignment
Free AI chatbots (ChatGPT free tier, etc)	£0	Unlimited	No diagnostic	No mark scheme
Kumon A-Level maths	£85-£120/mo	Twice weekly	None (worksheet-based)	None
Seneca Learning A-Level	£0-£10/mo	Unlimited	Topic level	Generic

For a Year 12 student targeting a B in AQA A-Level maths, the AI tutor covers the full syllabus, marks every past paper against the AQA mark scheme, identifies subtopic gaps to 3-decimal granularity, and costs £108 across the academic year. A private tutor covering the same syllabus in weekly 1-hour sessions costs £1,728 across the academic year and operates at topic level only.

When to start — the Y12 to Y13 maths transition plan

The peak buying window for an A-Level maths AI tutor is September to December of Year 13, when the student has finished the AS content and is preparing for the final A-Level exams in May-June. Starting earlier (June to August of Year 12, after the GCSEs finish) is even better — the AI can do the GCSE-to-A-Level bridge work over the summer and give the student a head start.

Here is the recommended timeline for a Year 12 student targeting an A or A* in A-Level maths by summer 2026:

• June to August 2025 (end of Year 11 / start of Year 12): AI does GCSE-to-A-Level bridge work. Topics: algebraic manipulation review, surds, indices, quadratics, simultaneous equations, trigonometry review, coordinate geometry review. Target: 80%+ fluency on Year 11 content within 4 weeks.
• September to December 2025 (Year 12 term 1): AI tracks schoolwork and drills AS content in parallel. Core topics: proof, algebra and functions, coordinate geometry, sequences and series, trigonometry, differentiation, integration, statistical sampling, probability, kinematics, forces. Target: 70%+ on AS mock by January.
• January to March 2026 (Year 12 term 2): AI consolidates AS content and starts Year 13 content. New topics: logarithms and exponentials, further differentiation and integration, numerical methods, vectors, statistical distributions, hypothesis testing, moments, work-energy. Target: 75%+ on Year 12 end-of-year exam.
• April to June 2026 (Year 12 term 3): AI runs full past paper practice. AQA Paper 1, Paper 2, Paper 3 from 2020, 2021, 2022, 2023, 2024, 2025. Mark every paper against the official mark scheme. Target: 75-80%+ on full past papers by end of Year 12.
• September to December 2026 (Year 13 term 1): AI runs full Year 13 content. New topics: parametric equations, implicit differentiation, integration techniques (parts, partial fractions, trig), differential equations, advanced mechanics topics, advanced statistics topics. Target: 80%+ on Year 13 mock by December.
• January to April 2027 (Year 13 term 2): AI runs intensive past paper practice. 20+ full past papers across AQA, OCR, OCR B (MEI), Edexcel. Timed conditions. Mark scheme grading. Target: 85%+ consistent by Easter.
• May to June 2027 (Year 13 final exams): AI runs final revision. Spaced repetition on all weak topics. Daily 60-90 minute sessions. Final mock exams in early May. Target: A or A* grade.

Internal links

This post is part of the A-Level cluster. Related content:

• AI Tutor for A-Level: 2026 UK Sixth-Form Parent Guide — the A-Level hub post covering all subjects
• AI Tutor for GCSE Maths: 2026 UK Parent Guide — for parents of Year 10-11 students preparing for GCSE
• AI Tutor for 11 Plus: 2026 UK Parent Guide — for parents of Year 5-6 students preparing for grammar school entry
• Parent GCSE AI Tutor 2026: Best UK Guide — for parents new to the AI tutor concept
• AI Tutor Pricing UK 2026: What Parents and Schools Actually Pay — for price-sensitive parents
• AI Tutor for A-Level Physics UK 2026 — mechanics, SHM, quantum physics, astrophysics

Sources

• AQA A-Level Mathematics Specification (7356, 7357) — aqa.org.uk
• OCR A-Level Mathematics A Specification (H230, H240) — ocr.org.uk
• OCR B (MEI) A-Level Mathematics Specification (H230, H240, H230/MEI) — ocr.org.uk
• Edexcel A-Level Mathematics Specification (9MA0) — pearson.com
• CIE A-Level Mathematics Specification (9709) — cambridgeinternational.org
• AQA A-Level Mathematics Past Papers and Mark Schemes 2018-2025 — aqa.org.uk
• Tutor Association 2026 UK Private Tuition Survey — tutorassociation.org.uk
• UCAS Tariff Tables 2026 — ucas.com
• Stanford University 2024 study on adaptive learning in mathematics — stanford.edu
• RAND Corporation 2024 meta-analysis on AI tutoring efficacy — rand.org

Related reading

• AI Tutor for A-Level: 2026 UK Sixth-Form Parent Guide — the full sixth-form picture
• AI Tutor for A-Level Chemistry — the most common STEM pair
• AI Tutor for A-Level Biology — for medicine applicants
• GCSE Maths Grade 9 Topics — the foundation under A-Level maths
• Year 6 SATs Parent Guide — where the maths pipeline starts

Related: A-Level Further Maths as the second A-Level

For UK sixth-form students targeting Oxbridge + engineering + mathematical sciences + computer science + economics + quantitative-finance offers, A-Level Further Maths is the second A-Level that gates the most prestigious university offers. Approximately 60 percent of A-Level Maths high-attainers in 2026 sit A-Level Further Maths alongside A-Level Maths. The OCR A + Edexcel + AQA + MEI specifications each have a different Core Pure + option-paper structure. See our <a href="/blog/ai-tutor-a-level-further-maths-uk-2026-playbook" class="text-[var(--brand-coral)] underline-offset-4 hover:underline font-semibold">AI tutor for A-Level Further Maths 2026</a> for the spec-routing + Core Pure + option-paper workflow + the 8 topics that cost the most marks.

Related: A-Level Physics as the second A-Level for Oxbridge engineering + natural sciences

For UK sixth-form students targeting Oxbridge + Imperial + UCL engineering + physics + natural sciences offers, A-Level Physics is the recommended second A-Level alongside A-Level Maths + A-Level Further Maths. The AQA + OCR A + Edexcel + CIE + WJEC A-Level Physics specifications each have a different paper structure + required-practical list + mathematical-requirements tier. The A-Level Physics content most relevant is the SHM + circular motion + gravitational fields + electric + magnetic fields + induction + quantum + nuclear + thermodynamics content. See our <a href="/blog/ai-tutor-a-level-physics-uk-2026-playbook" class="text-[var(--brand-coral)] underline-offset-4 hover:underline font-semibold">AI tutor for A-Level Physics 2026</a> for the spec-routing + required-practical + worked-model workflow.