Test series
Full exam-style papers, sat under real exam conditions.
Mock exams, past papers, predicted papers and topic tests. Marking combines instant auto-marking, an AI first pass and a tutor's final check, and results come with grades, topic mastery and analytics.
From building the paper to the grade
Built like the real exam
- 1
The admin builds the series. A series holds one or more papers, papers are split into sections, and sections hold questions from the bank.
Published with exam rules
- 2
The admin publishes it, now or on a scheduled date, and sets when it can be sat: a fixed sitting where everyone starts together, an open window, or any time.
Start, then read
- 3
The student opens Tests, picks the series and starts an attempt.
- 4
If the paper has reading time, the questions are visible but locked until it ends.
Answer against the clock
- 5
The student answers against a countdown; answers save automatically, and papers open one after another.
- 6
The student submits each paper, or all of them. There's a 30-second grace period after time runs out.
Marked in three layers
- 7
Objective answers are marked instantly, written answers get an AI first mark, and the tutor reviews, adjusts and finalises.
Grade, marks and mastery
- 8
Results are released, immediately or on a date the admin sets. The student sees their grade, marks per question and topic mastery.
Exam conditions, exam-grade marking
Kinds of test
Sat in the browser against the clock.
Download the paper, then upload photos of the written answers.
Some parts online, some written and uploaded.
The same rules for everyone
- Fixed sitting: everyone starts at the opening time. Late joiners are allowed up to a cutoff, and there can be a hard close.
- Open window: start any time between opening and closing.
- The clock is kept by the server, not the student's device. Students with an extra-time accommodation automatically get longer.
- A student who loses connection can always resume an attempt they've already started.
- Each series allows a set number of attempts, 3 by default.
For high-stakes papers
For high-stakes papers, proctoring requires fullscreen, blocks copy and paste, notices when the student switches tabs, takes webcam snapshots and checks the student's identity with a face match. Tutors can watch a live proctor view.
Instant, then AI, then the tutor
- Automatic: objective questions are marked instantly. For example, a multi-select question gives partial credit (correct picks minus wrong picks), numeric answers allow a tolerance, and text answers ignore capitals and spacing.
- AI first pass: written answers get a provisional mark from Veda.
- Tutor: the tutor checks each answer against the mark scheme and can adjust marks, write feedback, annotate, record audio comments and use an “AI suggest” button, then finalises. A paper with only objective questions finalises itself. A second marker and admin moderation keep marking consistent.
Feedback while practising
Students normally see results after marking. For practice-style series the admin can turn on instant feedback, so each answer is checked as it's given.
Releasing results
Results normally appear as soon as marking is done. For a mock exam, the admin can hold results back and release them on a set date, so the whole cohort gets them together.
Know exactly where you stand
IB grades come from fixed boundaries, and A-level and percentage scales are built in too. Students see “x% to the next grade”, their pass/fail against the pass mark, and a mastery score per topic. Tutors get a report per series. Admins get question-level analysis showing which questions are too easy, too hard or misleading. Parents see their child's results in the parent portal.
Question types
Over thirty types, from multiple choice and numeric to essays, graph sketching, code, Desmos and handwritten scan uploads.