About This Dataset
This dataset from Epoch AI provides a comprehensive record of notable AI models spanning decades of machine learning research and development. It tracks key metrics including model parameters, training compute (in FLOPs), training data size, and publication dates across diverse domains such as language processing, computer vision, multimodal systems, games, speech, and robotics. The data captures contributions from hundreds of organizations worldwide, offering valuable insights into the evolution of AI capabilities, the concentration of AI development among leading institutions, and the exponential growth trends in model scale and computational requirements that have defined the field's progress.
Key Metrics
Models by Domain
Top Organizations by Model Count
Included Files
notable_ai_models.csv
Notable AI models (976 records)
frontier_ai_models.csv
Frontier AI models data
large_scale_ai_models.csv
Large-scale AI models data
all_ai_models.csv
Complete AI models dataset
Data Table
| Model ↑ | Organization | Domain | Parameters | Date |
|---|---|---|---|---|
| (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) | NTT Communication Science Laboratories,Tohoku University | Language | 185.0M | 2018-08-30 |
| 3DDFA | Chinese Academy of Sciences,Michigan State University | Vision | 5.4M | 2015-11-23 |
| 3DMM-CNN | University of Southern California | Vision | 44.5M | 2016-12-15 |
| 4 layer QRNN (h=2500) | Salesforce Research | Language | 151.0M | 2018-03-22 |
| 6-Act Tether | Facebook AI Research,Georgia Institute of Technology | Robotics | 5.0M | 2021-08-03 |
Epoch AI, 'Parameter, Compute and Data Trends in Machine Learning'. Published online at epochai.org. Retrieved from https://epoch.ai/data/notable-ai-models [online resource]. Accessed 19 Jan 2026.