Epoch AI Data on AI Models
Project record
A public database of machine learning models maintained by Epoch AI, recording for each model estimates such as training compute, parameter count, training dataset size, training cost, training duration, and power draw. It is published as an interactive explorer and as CSV files covering notable models, large-scale models (trained with at least 10^23 FLOP), frontier models (top five in training compute when released), and all models.12
- Dataset hub: Data explorer and downloads (external site: epoch.ai)
- Documentation: Documentation (external site: epoch.ai)
- License: Creative Commons Attribution 4.0 license (external site: creativecommons.org)
Availability and license
Overall availability
Free to view in the web explorer and to download as CSV files (also bundled as ai_models.zip) without an account.1
Availability is separate from permission: read the license before using or redistributing.
Component reuse rights
- weights
- Unknown — no complete fact-level rights review
- code
- Unknown — no complete fact-level rights review
- data
- Reviewed qualifying license recorded — check scope and conditions
- documentation
- Unknown — no complete fact-level rights review
No complete system-rights review is recorded for this release.
Public materials checklist
| Item | Status | Notes and evidence |
|---|---|---|
| AccessCan the data be obtained, and on what terms? | Public | CSV downloads for each subset are linked from the dataset page.1 |
| ProvenanceAre the data's origins documented? | Partial | Epoch says models were collected from literature reviews, Papers With Code, historical accounts, highly cited publications, conference proceedings, and suggestions from individuals. The documentation includes a section on how values such as training compute are estimated.12 |
| DocumentationIs there a datasheet, card, or equivalent documentation? | Public | A documentation page defines the subsets and covers records, field definitions, and estimation methods, and a changelog is published.2 |
| LicensingAre the licensing terms stated? | Public | Creative Commons Attribution 4.0, stated on the dataset and documentation pages.12 |
| Stated limitationsDoes the documentation state known limitations or risks? | Partial | The dataset page says the list of models is not exhaustive and that the explorer only shows models for which Epoch has estimates, which are limited by publicly available information.1 |
What it is useful for
Studying trends in the resources used to train machine learning models over time, such as growth in training compute or parameter counts, with citations to the sources behind each estimate.1
Run and use notes
Organization context
U.S. eligibility
Eligible · basis: U.S.-governed project
The dataset is compiled and published by Epoch AI on its own site. Epoch Artificial Intelligence, Inc. gives a San Francisco, California address in its privacy notice and is listed in the IRS exempt-organization extract for California as a Section 501(c)(3) organization (see the Epoch AI record).134
Sources
This listing is not an endorsement, a safety assessment, or a federal approval.