ML Challenge

About

ML Challenge is a free, online machine-learning competition by Hack4Hack, open to students and working professionals from every country. Compete solo or in a team of up to 4, build models on real datasets across two 24-hour rounds, and win a share of ₹1,00,000 (₹1 lakh) in prizes, paid to the top 5 teams. Registration is free and closes on 2 December 2026, 11:59 PM IST.

Key facts

• What: an online machine-learning challenge (datathon) with a live leaderboard
• Who: college students from any year or branch, and working professionals, from any country
• Team size: 1 to 4 members (solo participants are welcome)
• Prize pool: ₹1,00,000 for the top 5 teams
• Registration: free, open now until 2 Dec 2026, 11:59 PM IST (18:29 UTC)
• Round 1: 3 Dec, 10:00 AM IST to 4 Dec, 10:00 AM IST (04:30 UTC to 04:30 UTC), free for every registered team
• Round 2: 8 Dec, 10:00 AM IST to 9 Dec, 10:00 AM IST (04:30 UTC to 04:30 UTC), for every team on the Round 1 leaderboard
• Round 2 fee: ₹299 per team for teams in India, or US$10 per team for teams outside India
• Results: 9 Dec 2026, 5:00 PM IST (11:30 UTC)
• Mode: fully online. All you need is a laptop and an internet connection

How it works

1. Register for free on Hack4Hack, solo or as a team of up to 4, by 2 Dec, 11:59 PM IST. Invite teammates by link or email.
2. The Round 1 problem statement and dataset go live on 3 Dec at 10:00 AM IST in your team workspace under Resources, with a sample submission and a data dictionary.
3. Round 1, the 24-hour ranking round (3 Dec, 10:00 AM to 4 Dec, 10:00 AM IST): train your model and upload predictions from the Leaderboard tab, up to 5 times a day (the limit resets at midnight IST). The leaderboard updates every 15 minutes.
4. Round 2 qualifiers on 5 Dec, 12:00 PM IST: every team on the Round 1 leaderboard (at least one scored upload) qualifies for Round 2 and is emailed. Teams confirm their place by paying the Round 2 fee (₹299 per team for teams in India, or $10 per team for teams outside India) from their workspace before Round 2 starts.
5. Round 2, 24 hours of live model development (8 Dec, 10:00 AM to 9 Dec, 10:00 AM IST): a new problem statement and a new dataset are released, with a fresh leaderboard, so every finalist starts from zero. Up to 10 uploads per team per day; files that fail validation do not count.
6. Submit your code (a GitHub repository or notebook link) from the Submission tab before Round 2 closes.
7. Final ranking and code verification on 9 Dec from 10:00 AM IST, on the private part of the Round 2 test set. Winners are announced on 9 Dec at 5:00 PM IST.

Prizes

• 1st place: ₹50,000
• 2nd place: ₹25,000
• 3rd place: ₹15,000
• 4th place: ₹5,000
• 5th place: ₹5,000
• Certificates for every team that uploads a scored submission, and certificates of achievement for the top 5

Fees

• Registering and Round 1 are free for everyone.
• Only teams going on to Round 2 pay a fee to confirm their place: ₹299 per team in India, or US$10 per team outside India. It is per team, not per person, and any member can pay online.

Who can participate

• Anyone, from any country: the challenge is fully online.
• College students from any year or branch, and working professionals.
• Solo or in a team of 2, 3 or 4 members. Each person can be in only one team.

Scoring

• The evaluation metric for each round is announced with its problem statement.
• Each round's leaderboard updates every 15 minutes and is scored on a public part of its test set.
• Each round's final ranking uses the hidden private part, based on each team's best public upload. Models that generalise win; overfitting the leaderboard does not.
• Every team on the Round 1 leaderboard qualifies for Round 2: there is no cutoff. Prizes are decided only by the Round 2 final ranking, on the new problem.

Submission format

The submission format for each round is announced with its problem statement, together with a sample submission file. Files that don't match the format are rejected straight away and don't use up an upload.

What you need

• A laptop and an internet connection
• Python and common libraries such as pandas, scikit-learn, XGBoost, LightGBM or PyTorch
• No paid GPU: free tiers such as Google Colab or Kaggle notebooks are enough

Why take part

• Win a share of ₹1 lakh and a certificate
• Practise on real datasets under real time pressure, with a live leaderboard
• Add a verified ranking and a GitHub project to your resume and portfolio

Prizes

• 1st place — ₹50K
• 2nd place — ₹25K
• 3rd place — ₹15K
• 4th place — ₹5K
• 5th place — ₹5K

Perks

• Free to register and play Round 1
• Certificate for every team that uploads a scored submission
• Certificate of achievement for the top 5
• Open to participants from every country

Timeline

1. Registration and team formation (7 Nov 2026 – 2 Dec 2026): Register for free, solo or as a team of up to 4 members, from any country. Invite teammates by link or email until registration closes on 2 Dec, 11:59 PM IST.
2. Round 1 problem statement and dataset release (3 Dec 2026 – 4 Dec 2026): The problem statement, training data, test file (without labels), a sample submission and the data dictionary are published in your team workspace under Resources.
3. Round 1: 24-hour ranking round (3 Dec 2026 – 4 Dec 2026): Runs for 24 hours, from 3 Dec, 10:00 AM to 4 Dec, 10:00 AM IST, starting as soon as the problem statement and dataset are released. Every registered team uploads predictions from the Leaderboard tab, up to 5 times a day. The leaderboard updates every 15 minutes, and every team that gets onto the Round 1 leaderboard qualifies for Round 2.
4. Round 2 qualifiers announced · Round 2 entry (5 Dec 2026 – 7 Dec 2026): Every team on the Round 1 leaderboard (at least one scored upload) qualifies for Round 2 and is emailed. Teams pay the Round 2 fee (₹299 per team for teams in India, or US$10 per team for teams outside India) from their workspace before Round 2 starts (8 Dec, 10:00 AM IST) to confirm their place.
5. Round 2: 24-hour live model development (8 Dec 2026 – 9 Dec 2026): A new problem statement and a new dataset are released to every Round 2 team at 10:00 AM IST on 8 Dec, with a fresh leaderboard. The round runs for 24 hours and closes at 10:00 AM IST on 9 Dec. Up to 10 uploads per team per day (the limit resets at midnight IST). No code submission is needed during the round: your ranking comes from your leaderboard uploads.
6. Final ranking and code verification (9 Dec 2026 – 9 Dec 2026): The Round 2 leaderboard switches to the private part of the test set. The top 5 teams are asked by email for their code, which is checked to make sure it reproduces their best upload.
7. Results (9 Dec 2026 – 9 Dec 2026): Winners announced at 5:00 PM IST on the hackathon page and by email.

Rules

1. Take part solo or in a team of up to 4 members. Each person can be in only one team, with one account. Participants from any country are welcome.
2. Train only on the dataset provided for each round. Open-source libraries and publicly available pretrained weights are allowed; outside labelled data for the same task and hand-labelling the test set are not.
3. Upload predictions in the submission format announced with each round's problem statement. A sample submission file comes with each dataset.
4. Round 1 runs for 24 hours, from 3 Dec, 10:00 AM to 4 Dec, 10:00 AM IST, starting right after the problem statement and dataset are released. It is open to every registered team, with up to 5 uploads a day. Every team that makes at least one scored upload, and so appears on the Round 1 leaderboard, qualifies for Round 2.
5. Teams that qualify confirm their Round 2 place by paying the entry fee (₹299 per team for teams in India, or US$10 per team for teams outside India) from their workspace before Round 2 starts (8 Dec, 10:00 AM IST).
6. Round 2 (8 Dec, 10:00 AM to 9 Dec, 10:00 AM IST, 24 hours) releases a new problem statement and a new dataset, with a new hidden test set. Each team gets up to 10 uploads a day (the limit resets at midnight IST); files that fail validation do not count.
7. Each round's evaluation metric is announced with its problem statement. Scores and the leaderboard update every 15 minutes, not instantly. Live leaderboards use a public part of the test set; final rankings use the hidden private part, based on each team’s best public upload. Prizes go by the Round 2 final ranking.
8. Before Round 2 closes, submit your code (a GitHub repository or notebook link) from the Submission tab. Prize-winning teams must be able to reproduce their best upload from that code.
9. Sharing code, predictions or labels between teams is not allowed and leads to disqualification.
10. The organizers’ decisions on qualification, verification and results are final.

ML Challenge: quick answers

What is ML Challenge?

ML Challenge is an online AI / ML datathon, organized by Hack4Hack. Free, online machine-learning challenge open to every country: two 24-hour rounds and ₹1 lakh in prizes. Solo or teams of up to 4. Students and working professionals. Register by 2 Dec 2026.

When is ML Challenge?

ML Challenge runs from 3 Dec 2026 to 9 Dec 2026, online. Registration is open until 2 Dec 2026. Stages: Registration and team formation (7 Nov 2026); Round 1 problem statement and dataset release (3 Dec 2026); Round 1: 24-hour ranking round (3 Dec 2026); Round 2 qualifiers announced · Round 2 entry (5 Dec 2026); Round 2: 24-hour live model development (8 Dec 2026); Final ranking and code verification (9 Dec 2026).

What is the last date to register for ML Challenge?

Registration for ML Challenge closes on 2 Dec 2026 (11:59 pm IST). Register on Hack4Hack before then; teammates can join with your team code.

What is the prize money at ML Challenge?

ML Challenge has ₹1,00,000 (₹1 lakh) in prizes. 1st place: ₹50,000, 2nd place: ₹25,000, 3rd place: ₹15,000, 4th place: ₹5,000, 5th place: ₹5,000.

Is ML Challenge free?

Yes, registering for ML Challenge is free. Teams shortlisted for the next round confirm their place with a ₹299 fee per team.

Who can participate in ML Challenge?

College students and working professionals can take part, in teams of 1 to 4. It is fully online, so you can join from anywhere.

Can I join ML Challenge from outside India?

Yes. ML Challenge runs fully online, so you can take part from any country unless its rules say otherwise. It starts 3 Dec, 10:00 am IST (3 Dec, 4:30 am UTC).

Frequently asked questions

What is ML Challenge?

ML Challenge is a free, online machine-learning competition by Hack4Hack with ₹1,00,000 in prizes. Teams of 1 to 4 build models on real datasets in two 24-hour rounds (3–4 Dec and 8–9 Dec 2026), ranked on a live leaderboard.

Is ML Challenge free?

Yes. Registering and Round 1 are free for everyone. Only teams that go on to Round 2 pay a fee to confirm their place: ₹299 per team in India, or US$10 per team outside India (per team, not per person).

Can international students and professionals take part?

Yes. ML Challenge is fully online and open to participants from every country. Teams outside India pay a Round 2 fee of US$10 per team if they go on to Round 2; registration and Round 1 are free.

Who can participate?

College students from any year or branch and working professionals, from any country.

Can I take part alone?

Yes. Teams can have 1 to 4 members, so you can compete solo. Solo participants are fully eligible for Round 2 and the prizes.

What is the last date to register?

Registration closes on 2 December 2026 at 11:59 PM IST (18:29 UTC).

What are the prizes?

₹1,00,000 in total for the top 5 teams: 1st place ₹50,000, 2nd ₹25,000, 3rd ₹15,000, 4th ₹5,000 and 5th ₹5,000. Prizes are decided by the Round 2 final ranking.

What is the timeline?

Registration until 2 Dec, 11:59 PM IST. Round 1: 3 Dec, 10:00 AM to 4 Dec, 10:00 AM IST. Round 2 qualifiers announced: 5 Dec, 12:00 PM IST. Round 2: 8 Dec, 10:00 AM to 9 Dec, 10:00 AM IST. Winners: 9 Dec, 5:00 PM IST.

When is the problem statement revealed?

Round 1’s problem statement and dataset go live on 3 Dec at 10:00 AM IST (04:30 UTC) in your team workspace under Resources, and the 24-hour Round 1 opens at the same time. Round 2 has a new problem statement and a new dataset, released to every Round 2 team on 8 Dec at 10:00 AM IST.

Who qualifies for Round 2?

Every team on the Round 1 leaderboard: make at least one scored upload during Round 1 (3 Dec, 10 AM to 4 Dec, 10 AM IST) and your team qualifies for Round 2. There is no ranking cutoff. Qualified teams are emailed on 5 Dec.

How do we pay the Round 2 fee?

From the Leaderboard tab of your team workspace, online by UPI, card or netbanking (international cards are accepted), before Round 2 starts on 8 Dec at 10:00 AM IST. Any team member can pay.

What exactly do we upload?

A file of your predictions for the test set. The exact submission format is announced with each round’s problem statement, together with a sample submission file.

How is it scored?

The evaluation metric for each round is announced with its problem statement. Live leaderboards use a public part of the test set; final rankings use the hidden private part, which rewards models that generalise.

How many times can we upload?

Up to 5 times per team per day in Round 1, and up to 10 times per team per day in Round 2. The limit resets at midnight IST, and uploads that fail validation do not use it up.

When do we see our score?

The leaderboard and your own scores update every 15 minutes from the start of each round, and the final ranking comes right after the round closes. Files with format errors are rejected right away and do not use up an upload.

Do I need a GPU?

No. Free tiers such as Google Colab or Kaggle notebooks are enough.

Which languages and libraries can I use?

Any. Most teams use Python with pandas, scikit-learn, XGBoost, LightGBM or PyTorch. Open-source libraries and publicly available pretrained weights are allowed; outside labelled data for the same task is not.

Will I get a certificate?

Yes. Every team that uploads a scored submission gets a participation certificate, and the top 5 teams get certificates of achievement.