The AI Museum
AI Analytics1,072 recordsRecounted 23 September 2026

Frontier language models from United States of America

38

Models

17 organisations

29%

Open weights

9 open, 22 closed

38

Frontier

By Epoch's flag

1.6T

Largest traceable

Switch

1959–2025

Span

20 with a traceable size

Records

38
Frontier language models from United States of America, newest first. Each row links to the source Epoch AI records for it.
ModelOrganisationReleasedParametersWeights
Grok 4xAI2025-073.0T *Closed
GPT-4.5OpenAI2025-02Closed
Grok 3xAI2025-023.0T *Closed
Grok-2xAI2024-08Closed
Llama 3.1-405BMeta AI2024-07405BOpen
Nemotron-4 340BNVIDIA2024-06340BOpen
Gemini 1.5 ProGoogle DeepMind2024-02Closed
Inflection-2Inflection AI2023-11Closed
Amazon TitanAmazon2023-09200B *Closed
Claude 2Anthropic2023-07Closed
PaLM 2Google2023-05340B *Closed
Flan-PaLM 540BGoogle2022-10540BClosed
U-PaLM (540B)Google2022-10540BClosed
Minerva (540B)Google2022-06540BClosed
PaLM (540B)Google Research2022-04540BClosed
GPT-3.5 (davinci-002)OpenAI2022-03Closed
Megatron-Turing NLG 530BMicrosoft2021-10530BClosed
FLAN 137BGoogle Research2021-09137BClosed
SwitchGoogle2021-011.6TOpen
mT5-XXLGoogle2020-1013BOpen
GShard (dense)Google2020-062BClosed
GPT-3 175B (davinci)OpenAI2020-05175BClosed
MeenaGoogle Brain2020-013BClosed
XLM-RoBERTaFacebook AI2019-11550MOpen
T5-11BGoogle2019-1011BOpen
T5-3BGoogle2019-103BOpen
Megatron-BERTNVIDIA2019-094BClosed
Megatron-LM (8.3B)NVIDIA2019-098B *Closed
RoBERTa LargeFacebook2019-07355MOpen
XLNetCarnegie Mellon University (CMU)2019-06340MOpen
GNMTGoogle2016-09278M *Closed
SNM-skipGoogle2014-1262B *Not recorded
Seq2Seq LSTMGoogle2014-092BNot recorded
DistBelief NNLMGoogle2013-01Not recorded
RNN LMJohns Hopkins University2010-0970M *Not recorded
KN-LMGoogle2007-0621B *Not recorded
SB-LMGoogle2007-06300B *Not recorded
Pandemonium (morse)Massachusetts Institute of Technology (MIT)1959-02Not recorded

An asterisk marks a parameter figure the source does not consider traceable to the people who built the model. That judgement is the source’s and it has false negatives: Kimi K3 is marked speculative at 2.8T while Moonshot’s own model card states the figure. Why that distinction carries the whole exhibit.

Counts are of models Epoch AI records as notable, not of every model released. The most recent month is always incomplete, because a model enters the dataset when it is reviewed rather than when it ships.

Epoch AI, 'Data on AI Models'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/data/ai-models-documentation' Used under CC BY 4.0.

How this museum is made