Australian National Olympiad in AI · NOAI 2026

Discovering Australia's brightest young AI talent.

Open to every Australian secondary student in Years 7–12. No prior AI or programming experience is needed for Round 1 — the pathway takes you from a 45-minute online test to the International Olympiad in Artificial Intelligence.

New here? Start here How to prepare

Team Australia at IOAI 2026 in Astana, Kazakhstan

🥉 Bronze + 2 Honourable Mentions · Astana 2026

IOAI editions contested

Gold

Practice Round award · Bulgaria 2024

Bronze

Theory Round medal · Beijing 2025

🥉 + 2

Bronze and two Honourable Mentions · Astana 2026

Team Australia

Three editions, rising every year.

Read the story of each Australian delegation to the International Olympiad in Artificial Intelligence. IOAI 2026 in Astana brought together 466 contestants from 108 countries and territories. IOAI 2027 will be held at Nanyang Technological University, Singapore, 4–10 July 2027.

Team Australia in Astana, Kazakhstan, 2026

2026 · Astana, Kazakhstan

One Bronze Medal and two Honourable Mentions — a major step forward.

2025 · Beijing, China

Australia's first Theory Round Bronze Medal.

Team Australia in Burgas, Bulgaria, 2024

2024 · Burgas, Bulgaria

Debut edition — Gold Award in the Practice Round.

AUScholarship · National AI Talent Pathway

We are inviting 200 Australian schools.

One funded place each — and the door is open to the rest. A full term of frontier AI — taught online, built around asking better questions, and priced so that no school is left out.

Nine Australian students still in middle school already hold medals or awards from international AI olympiads — won in fields of senior secondary and first-year university competitors. AUScholarship opens the pathway that produced them to one student from each of 200 schools.

AUScholarship national AI talent pathway

2026 Australian National Olympiad in AI

The road to IOAI 2026.

Eligibility: open to all Australian secondary school students in Years 7–12 with an interest in artificial intelligence, programming, and problem-solving.

1
NOAI Level Assessment — Round 1

Registration 8 Nov 2025 – 27 Mar 2026 · Assessment 29 Mar 2026 (late window 5 Apr 2026)
- Online written test · 45-minute multiple-choice format (A–E options) · no programming.
- Assesses fundamental understanding of AI, mathematics and logical reasoning.
- All students from Round 1 can participate in Round 2.
2
NOAI Advanced AI and Theory Test — Round 2

26 April 2026 · online, 90 minutes, 100 marks

Mixed paper: multiple-choice, short answer and Python programming on the browser-based platform at exam.ioai.org.au. Closed book, internet disabled, Python interpreter available in the programming section.
Exam environment & preparation guide

Published so that every candidate has identical information about the exam platform, permitted software, and preparation direction. No topic-level content is disclosed.

Programming environment (server side)

The exam server is pre-configured — candidates do not need to install anything.
- Python3.13.x
- NumPy2.2.x
- pandas2.2.x
- scikit-learn1.6.x
- Operating systemLinux x86_64
✔ Permitted imports: numpy, pandas, scikit-learn, and the Python standard library.
✖ Not permitted: PyTorch, TensorFlow, Keras, JAX, scipy.signal, and any other deep-learning or signal-processing framework; external dataset downloads; any form of network request.

Recommended local practice setup
```
pip install "numpy>=1.21" "pandas>=1.3" "scikit-learn>=1.0"
```
Python 3.9 or newer is recommended. You do not need to match the server's exact minor version — write code against the stable public API of NumPy / pandas / scikit-learn.

Preparation guidance (direction, not scope)
- Mathematical foundations — matrix and vector operations, probability and expectation, basic calculus.
- Core ML literacy — read and reason about common AI concepts; perform basic error-analysis calculations by hand.
- Python fluency — NumPy array construction, slicing, np.zeros, np.sum, np.mean, np.median, broadcasting; pandas read_csv, fillna, row filtering, exporting to CSV; scikit-learn train_test_split, .fit / .predict, standard metrics.
- Coding style — the exam marks emphasise runnable code. Get it working first, polish second; partial credit is given for correct logic.
On the day
- Log in to exam.ioai.org.au at least 15 minutes before the start and complete the three-layer invigilation check: locked browser + screen recording + Zoom.
- Programming questions include an in-browser code editor; you can run and re-run your code before submitting. Your most recent submission is saved automatically when the exam ends.
3

NOAI Australia — Final Round Selection (APOAI stage)

13 June 2026 · online — no travel to Sydney required

Eligibility: participants qualified from Round 1 + Round 2. The Final Round adopts APOAI (Asia-Pacific Olympiad in Artificial Intelligence) competition standards — consistent with the IOAI selection ecosystem — with a unified leaderboard, GPU-enabled environment (e.g. Bohrium/Kaggle) and automatic evaluation, used across participating Asia-Pacific regions.

Results from the APOAI Final Round are the primary selection basis for the NOAI Australia National Team.

APOAI 2026 was hosted by China on 13 June 2026 in a distributed format, with 126 contestants from 18 countries and regions; Australia's eight contestants took one Silver, one Bronze and two Honourable Mentions. APOAI 2027: Moscow, Russia, April 2027. Details at apoai.org.
🏆

IOAI International Round — Astana, Kazakhstan

2 – 8 August 2026 · Australia National Team (2 teams, 8 members)

Hybrid format (online + on-site training and competition): IOAI individual challenge, team challenge, and collaborative workshops.

Read how Team Australia went in Astana →

Next: IOAI 2027 — Nanyang Technological University, Singapore, 4–10 July 2027.