What is Kaggle?
Kaggle currently prioritizes independent, community-driven AI evaluations, integrating competitions, hackathons, and benchmarks to help researchers and practitioners compare methods under well-defined problem statements and evaluation criteria. The homepage features sections for Benchmarks, the Game Arena, and data resources.
The platform provides datasets, models, public Notebooks, and evaluation tasks, allowing you to review others’ methods and results before conducting your own experiments. The benchmark creation process supports custom tasks, private test sets, and reproducible leaderboards; please refer to the relevant documentation for specific participation requirements.
When using the service, please first verify the task rules, data permissions, submission methods, and available computing resources before deciding whether to participate. The leaderboard and prize money do not guarantee individual earnings, and publicly available community resources do not mean that all data and models can be freely redistributed for commercial purposes.
Key features
AI Competitions and Hackathons
Open benchmarks and leaderboards
Datasets and Model Resources
Public Notebook and Experimental Methods
Custom Evaluations and Community Documentation
Use cases and audience
Data science learners, AI researchers, and model evaluation teams
- AI Evaluation and Method Comparison
- Data Science Learning and Practice
- Reproducible Experiments and Competition Exploration
Tasks, data, models, and experimental methods
Experimental code, evaluation results, leaderboards, and community practice content
How to use
- 01
Choose the appropriate benchmark, competition, or data topic.
- 02
Read the rules, licenses, and existing methods and Notebooks.
- 03
Conduct experiments according to the documentation, verify the results, and meet the submission requirements.
Official website & sources
Sources
Updated on 2026-10-24
