While IMO 2026 runs in Shanghai, many competition pages still offer PDF downloads and a promise to add full statements later. Here is how we built the full IMO archive on problems.cc.
A student preparing for the International Mathematical Olympiad should not have to download a PDF, zoom in on a scan, and hope the notation survived the export.
That is still what many competition pages offer. The official problem sets are available to download. The full statements on the page are "coming soon". If you happen to have the English LaTeX source, please get in touch.
Useful as an archive. Useless as a daily practice system.
While IMO 2026 runs in Shanghai (10-21 July; contest days 15-16 July), we had a practical problem at Exact Science: our olympiad students needed every IMO statement in the same structured form as the UKMT and BMO papers they already work through on problems.cc. The official ecosystem was still catching up. So we built the archive ourselves.
This is what we did, and why it mattered.
The gap we were filling
Olympiad preparation is not "collect every past paper". It is deliberate problem-solving on problems that match the competition's habits.
A PDF bundle gets in the way of that:
- you cannot jump to question 4 of 2019 without scrolling
- you cannot assign one paper to a student and know they saw the same wording you did
- you cannot build a lesson around a single statement when the statement lives inside a fifty-page scan
- you cannot track which years a student has actually attempted
The IMO event page on olympiads.co.uk reflects the wider picture: valuable links, downloadable sets, and an honest note that full statements are still being added page by page. That is fine for a competition directory. It is not enough when you are coaching someone who needs six proof problems under IMO conditions, week after week.
We teach olympiad-style maths. Our students needed the problems digitalised in a form we could teach from.
What we put on problems.cc
We ingested the historical IMO archive onto problems.cc: every paper from 1959 to 2026, each with six problems, Day 1 (questions 1-3) and Day 2 (questions 4-6).
That is decades of the hardest school-level mathematics in the world, now browsable as structured papers rather than a folder of downloads.
What is there: problem statements, year by year, in a consistent layout.
What this is not: a claim that every solution in the literature is bundled in. The point is access to the tasks themselves, so students can attempt them properly.
If the IMO would find these structured statements useful, we are happy to share them; they are not yet on the official IMO site.
How we built the archive
We did not hand-type six hundred problems from memory. We also did not run a black box and publish whatever came out.
The workflow was deliberately conservative:
- Start from official sources. The IMO's own publications and problem sets, plus the historical PDFs where that was the only public format.
- Extract structured statements. Each problem as its own object, with LaTeX preserved where the source allowed it.
- Validate before publish. Broken notation, truncated diagrams, or mismatched numbering get caught before a student sees them.
- Human review on awkward years. Older scans and odd layouts still need a mathematician's eye. Some editions are messier than others. We would rather delay one paper than ship a garbled statement.
That is the same philosophy we use elsewhere at Exact Science: treat preparation as a research problem, then turn the output into something a student can actually use.
The mistake I see in IMO preparation
A strong BMO student opens a random IMO paper from the 1990s, gets stuck on question 2 in twenty minutes, and concludes they are "not ready".
Often they are ready for a different question from the same paper. Or for an easier year. Or for one problem attempted properly instead of six skimmed.
Without a usable archive, students default to folklore: "2011 was hard", "geometry years are scary", "start with 2004". None of that is a training plan.
With papers on problems.cc, a coach can say: attempt 2016 Day 1 under contest time, then we review whether your block was technique or panic. That is a different conversation.
How Exact Science fits
We run olympiad maths teaching for students on the UK pathway: UKMT challenges, BMO rounds, camps, and the long climb towards international selection.
problems.cc is where that teaching lands in practice. Past papers, curated sets, daily puzzles, and now the full IMO corpus live on the same platform our students already use. Lessons point to a URL. Homework is a specific paper, not "find something online".
The IMO is not the first competition our students meet. For most, it is a horizon. But the problems are worth studying long before Shanghai is realistic. They train proof length, idea generation under no template, and the discipline of writing mathematics clearly. Those habits show up in BMO Round 2 and in every serious written paper.
A simple protocol for students
If you want to use the archive well, not just admire it:
- Pick one paper, not a decade. Choose a year, open Day 1 or Day 2, and treat it as a mock (4.5 hours, two problems at a time if you are imitating IMO conditions).
- Attempt before you read. No solution hunting after ten minutes of discomfort. Write what you tried.
- Review with structure. What was the entry point? Where did you stall? Was the gap knowledge or courage?
- Log the year and question numbers. So next month you are not accidentally repeating the same paper while thinking you are "doing IMO prep".
- Step back to your level when needed. BMO and national olympiad papers are the right floor for most UK students. IMO papers are the ceiling you reach for deliberately, not every Tuesday.
Why this matters while IMO 2026 is live
The competition in Shanghai is the headline. For students watching from the UK, it is also a reminder of what the top of the subject looks like: six problems, two days, no partial credit for a clever idea you could not finish.
You do not need to be in China to learn from that standard. You need the problems in front of you, regularly, with feedback.
We digitalised the archive because our students deserved that. If you are serious about olympiad maths, browse the papers at problems.cc/papers/international-mathematical-olympiad. Pick one year. Attempt one problem properly. See what changes when the statement is the starting point, not the PDF hunt.
And if you want teaching that matches that standard, Exact Science is where we do the rest of the work.