Data on AI Models | Epoch AI
https://epoch.ai/data/ai-models
Epoch AI model database — independent longitudinal data on model releases and compute, useful for corroborating vendor claims about lineage and scale
3
Lines added
3
Lines removed
Was read
2026-09-24
d0dad2581533d1f2…
Changed by
2026-09-25
5bf8be14fd31ba66…
3 lines added, 3 lines removed
Left column is the version read 2026-09-24, right is 2026-09-25. Unchanged stretches are elided. Both bodies are kept in full and are addressed by the hashes above.
⋯ 104 unchanged lines
105105 Citations
106106 Epoch AI’s data is free to use, distribute, and reproduce provided the source and authors are credited under the Creative Commons Attribution license .
107107
108108 Citation
109−Epoch AI, 'Data on AI Models'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/data/ai-models' [online resource]. Accessed 24 Sep 2026.
109+Epoch AI, 'Data on AI Models'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/data/ai-models' [online resource]. Accessed 25 Sep 2026.
110110
111111 BibTeX Citation
112112 @misc{EpochAIModels2025,
113113 title = {{Data on AI Models}},
114114 author = {{Epoch AI}},
115115 year = {2026},
116116 month = {9},
117117 url = {https://epoch.ai/data/ai-models},
118−note = {Accessed: 24 Sep 2026}
118+note = {Accessed: 25 Sep 2026}
119119 }
120120
121121 Python Import
122122 import pandas as pd
⋯ 12 unchanged lines
135135 What are notable, frontier, and large-scale models?
136136
137137 We flag models as notable if they advanced the state of the art, achieved many citations in an academic publication, had over a million monthly users, were highly significant historically, or were developed at a cost of over one million dollars. You can learn more about these notability criteria by reading our AI Models Documentation .
138138
139−Frontier models are models that were in the top 10 by training compute at the time of their release, a threshold that grows over time as larger models are developed.
139+Frontier models are models that were in the top 5 by training compute at the time of their release, a threshold that grows over time as larger models are developed.
140140
141141 Large-scale models are models that were trained with over 10^23 FLOP of compute, which is a static threshold that is used in some AI regulatory frameworks.
142142
143143 Why are the number of models in the database and the results in the explorer different?