From Google’s NASv3 RL network, trained on 800 GPUs in 2016, to Meta’s Llama 3.1 405B, using 16,384 H100 GPUs in 2024, the number of processors used increased by a factor of over 20. Gemini Ultra was trained with an even larger number of TPUs, but precise details were not reported.
| Series | Model | Organization | Publication date | Training compute (FLOP) | Hardware quantity |
|---|---|---|---|---|---|
| Top 3 | KN-LM | 2007-06-22 | 7.7e+17 | 400 | |
| Top 3 | SB-LM | 2007-06-22 | 1.4e+18 | 1,500 | |
| Top 3 | High Performance CNN (NORB) | IDSIA, SUPSI | 2011-07-16 | 2.6e+16 | 4 |
| Top 3 | DistBelief NNLM | 2013-01-16 | 2.6e+18 | 180 | |
| Top 3 | NASv3 (CIFAR-10) | Google Brain | 2016-11-05 | 2.2e+21 | 800 |
| Top 3 | AlphaZero | DeepMind | 2017-12-05 | 1.1e+20 | 5,064 |
| Top 3 | RoBERTa Large | Facebook, University of Washington | 2019-07-01 | 8.5e+21 | 1,024 |
| Top 3 | OpenAI Five | OpenAI | 2019-12-13 | 6.7e+22 | 1,536 |
| Top 3 | GPT-3 175B (davinci) | OpenAI | 2020-05-28 | 3.1e+23 | 10,000 |
| Top 3 | ViT-G/14 | Google Brain, Google Research | 2021-06-08 | 5.8e+22 | 2,048 |
| Top 3 | Megatron-Turing NLG 530B | Microsoft, NVIDIA | 2021-10-11 | 8.6e+23 | 4,480 |
| Top 3 | PaLM (540B) | Google Research | 2022-04-04 | 2.5e+24 | 6,144 |
| Top 3 | GPT-4 | OpenAI | 2023-03-15 | 2.1e+25 | 25,000 |
| Top 3 | Amazon Titan | Amazon | 2023-09-28 | 4.8e+24 | 13,760 |
| Top 3 | Gemini 1.0 Ultra | Google DeepMind | 2023-12-06 | 5.0e+25 | 57,000 |
| Top 3 | Llama 3.1-405B | Meta AI | 2024-07-23 | 3.8e+25 | 16,384 |
| Top 3 | Grok 3 | xAI | 2025-02-17 | 3.5e+26 | 80,000 |
| Top 3 | Llama 4 Behemoth (preview) | Meta AI | 2025-04-05 | 5.2e+25 | 32,000 |
| Top 3 | Grok 4 | xAI | 2025-07-09 | 5.0e+26 | 200,000 |
| Other models | Theseus | Bell Laboratories | 1950-07-02 | 4.0e+1 | NA |
| Other models | SNARC | Harvard University | 1952-01-08 | NA | NA |
| Other models | Genetic algorithm | Institute for Advanced Study | 1954-07-02 | NA | NA |
| Other models | Self Organizing System | Massachusetts Institute of Technology (MIT) | 1955-03-01 | NA | NA |
| Other models | Sequence-based pattern recognition | Massachusetts Institute of Technology (MIT) | 1955-03-01 | NA | NA |
| Other models | Perceptron Mark I | 1957-01-01 | 6.9e+5 | NA | |
| Other models | Pandemonium (morse) | Massachusetts Institute of Technology (MIT) | 1959-02-01 | 6.0e+8 | NA |
| Other models | Samuel Neural Checkers | IBM | 1959-07-01 | 4.3e+8 | NA |
| Other models | Pattern recognition and reading by machine | Sandia Corporation | 1959-12-01 | NA | NA |
| Other models | Perceptron (1960) | Cornell Aeronautical Laboratory | 1960-03-30 | 7.2e+8 | NA |
| Other models | ADALINE | Stanford University | 1960-06-30 | 6.6e+3 | NA |
| Other models | Linear Decision Functions | Bell Laboratories | 1962-06-01 | 1.6e+6 | NA |
| Other models | MADALINE I | Stanford University | 1962-07-01 | NA | NA |
| Other models | Print Recognition Logic | IBM | 1963-01-01 | 2.2e+7 | NA |
| Other models | Heuristic Reinforcement Learning | Purdue University | 1965-10-01 | 1.1e+6 | NA |
| Other models | LTE speaker verification system | IBM | 1966-11-01 | 1.1e+8 | NA |
| Other models | GLEE | University of Edinburgh | 1968-07-01 | NA | NA |
| Other models | Boxes (pole) | University of Edinburgh | 1968-07-01 | NA | NA |
| Other models | Decision tree adaline | Tokyo Medical and Dental University | 1969-05-01 | NA | NA |
| Other models | Piecewise linear model | University of Kansas | 1973-11-01 | NA | NA |
| Other models | Cognitron | Biological Cybernetics | 1975-09-01 | 5.2e+6 | NA |
| Other models | Neocognitron | 1980-04-01 | 2.7e+8 | NA | |
| Other models | Kohonen network | Helsinki University of Technology | 1981-07-25 | NA | NA |
| Other models | Hopfield network | California Institute of Technology | 1982-04-01 | NA | NA |
| Other models | ASE+ACE | University of Massachusetts Amherst | 1983-09-01 | 3.2e+8 | NA |
| Other models | Hierarchical Cognitron | 1984-04-01 | NA | NA | |
| Other models | Error Propagation | 1986-01-03 | NA | NA | |
| Other models | Distributed representation NN | Carnegie Mellon University (CMU) | 1986-08-15 | 3.9e+8 | NA |
| Other models | Back-propagation | 1986-10-01 | 6.7e+8 | NA | |
| Other models | Optimized Multi-Scale Edge Detection | Massachusetts Institute of Technology (MIT) | 1986-11-01 | NA | NA |
| Other models | NetTalk (transcription) | Princeton University | 1987-06-06 | 2.8e+10 | NA |
| Other models | NetTalk (dictionary) | Princeton University | 1987-06-06 | 2.8e+10 | NA |
| Other models | Translation-invariant MLP | Carnegie Mellon University (CMU) | 1987-06-15 | 1.8e+10 | NA |
| Other models | MLN-ASR | McGill University | 1988-08-01 | 3.0e+8 | NA |
| Other models | Q-learning | University of London | 1989-01-01 | NA | NA |
| Other models | MLP baggage detector | 1989-01-01 | NA | NA | |
| Other models | Truck backer-upper | Stanford University | 1989-06-18 | NA | NA |
| Other models | Invariant image recognition | Complutense University of Madrid | 1989-06-18 | 2.7e+10 | NA |
| Other models | Speaker-independent vowel classification | University of Washington | 1989-11-27 | 7.5e+9 | NA |
| Other models | Handwritten Digit Recognition System | AT&T | 1989-11-27 | 1.8e+11 | NA |
| Other models | Zip CNN | AT&T, Bell Laboratories | 1989-12-01 | 1.5e+12 | NA |
| Other models | NETtalk reimplementation | Oregon State University | 1990-06-01 | 3.6e+10 | NA |
| Other models | Bankruptcy-NN | NA | 1990-06-17 | 3.1e+9 | NA |
| Other models | SexNet compression | NA | 1990-10-01 | 7.9e+10 | NA |
| Other models | SexNet classification | NA | 1990-10-01 | NA | NA |
| Other models | ISR network | Stanford University | 1990-10-01 | NA | NA |
| Other models | RAAM | NA | 1990-11-01 | NA | NA |
| Other models | Weight Decay | NA | 1991-12-02 | 7.5e+10 | NA |
| Other models | TD-Gammon | IBM | 1992-05-01 | 1.8e+13 | NA |
| Other models | Golem | Alan Turing Institute | 1992-10-01 | NA | NA |
| Other models | Cancer drug mechanism prediction | National Cancer Institute | 1992-10-16 | 5.3e+7 | NA |
| Other models | Boosting | Bell Laboratories | 1992-11-30 | NA | NA |
| Other models | IBM-5 | IBM | 1993-06-15 | NA | NA |
| Other models | Siamese-TDNN | Bell Laboratories | 1993-08-01 | 1.3e+13 | NA |
| Other models | ANN Eye Tracker | NA | 1993-11-29 | 1.8e+10 | NA |
| Other models | Ceramic-MLP | Sapienza Università di Roma | 1994-01-07 | 4.5e+9 | NA |
| Other models | GroupLens | Massachusetts Institute of Technology (MIT) | 1994-10-22 | NA | NA |
| Other models | Predictive Coding NN | Technical University of Munich | 1994-12-02 | 1.9e+13 | NA |
| Other models | NeuroChess | NA | 1994-12-02 | 8.6e+11 | NA |
| Other models | Mixture of linear models | NA | 1994-12-02 | 4.5e+11 | NA |
| Other models | JPMAX | NA | 1994-12-02 | 8.1e+7 | NA |
| Other models | Support Vector Machines | AT&T, Bell Laboratories | 1995-09-01 | NA | NA |
| Other models | LISSOM | University of Texas at Austin | 1995-11-27 | 2.0e+11 | NA |
| Other models | MUSIC perceptron | NA | 1996-06-03 | 8.8e+11 | NA |
| Other models | System 11 | Carnegie Mellon University (CMU) | 1996-06-18 | 2.6e+10 | NA |
| Other models | AdaBoost.M2 Digit Recognition | AT&T | 1996-07-03 | NA | NA |
| Other models | SOM-CNN | NA | 1997-01-31 | 3.1e+10 | NA |
| Other models | Deep Blue | IBM | 1997-05-01 | NA | NA |
| Other models | Bidirectional RNN | 1997-11-01 | NA | NA | |
| Other models | LSTM | Technical University of Munich | 1997-11-15 | 3.2e+13 | NA |
| Other models | LeNet-5 | AT&T | 1998-11-01 | 2.8e+12 | NA |
| Other models | LSTM with forget gates | IDSIA | 1999-01-02 | NA | NA |
| Other models | RECONTRA-categorized | NA | 1999-06-02 | 8.0e+12 | NA |
| Other models | RECONTRA-uncategorized | NA | 1999-06-02 | 3.9e+12 | NA |
| Other models | IBM Model 4 | 1999-07-02 | NA | NA | |
| Other models | Neural LM | 2000-11-28 | 6.3e+15 | NA | |
| Other models | PoE MNIST | University College London (UCL) | 2000-11-28 | 5.2e+13 | NA |
| Other models | Immediate trihead | Brown University | 2001-07-06 | NA | NA |
| Other models | Gradient Boosting Machine | Stanford University | 2001-10-01 | NA | NA |
| Other models | Decision tree (classification) | Mitsubishi Electric Research Labs, Compaq CRL | 2001-12-08 | 6.3e+13 | NA |
| Other models | Thumbs Up? | Cornell University, IBM | 2002-05-28 | NA | NA |
| Other models | Statistical Shape Constellations | California Institute of Technology | 2003-01-01 | NA | NA |
| Other models | LDA | Stanford University | 2003-02-02 | NA | NA |
| Other models | NPLM (Brown) | 2003-03-15 | 1.3e+14 | NA | |
| Other models | NPLM (AP News) | 2003-03-15 | 1.7e+15 | NA | |
| Other models | Invariant CNN | New York University (NYU) | 2004-06-27 | 9.7e+11 | NA |
| Other models | LMICA | NA | 2004-12-01 | 2.8e+15 | NA |
| Other models | Hierarchical LM | NA | 2005-01-06 | 1.2e+14 | 1 |
| Other models | Histograms of Oriented Gradients | INRIA | 2005-06-25 | NA | NA |
| Other models | RankNet | Microsoft Research, Microsoft | 2005-08-07 | 3.5e+12 | NA |
| Other models | TFE SVM | 2006-02-02 | NA | NA | |
| Other models | FAST | University of Cambridge | 2006-05-07 | NA | NA |
| Other models | Spatial Pyramid Matching | 2006-06-17 | NA | NA | |
| Other models | SVM-CNN | New York University (NYU) | 2006-06-17 | 7.5e+14 | 1 |
| Other models | Dimensionality Reduction | University of Toronto | 2006-07-18 | NA | NA |
| Other models | Deep Belief Nets | 2006-07-18 | NA | NA | |
| Other models | Local Binary Patterns for facial recognition | University of Oulu, IEEE | 2006-12-01 | NA | NA |
| Other models | Greedy layer-wise DNN training | 2006-12-04 | NA | NA | |
| Other models | λ-WASP | UT Austin | 2007-06-01 | NA | NA |
| Other models | BLSTM for handwriting (1) | 2007-09-23 | NA | NA | |
| Other models | Enhanced Neighborhood-Based Filtering | AT&T | 2007-10-28 | NA | NA |
| Other models | BLSTM for handwriting (2) | 2007-12-03 | NA | NA | |
| Other models | Deep Multitask NLP Network | NEC Laboratories | 2008-07-05 | NA | NA |
| Other models | Denoising Autoencoders | 2008-07-05 | NA | NA | |
| Other models | Sparse digit recognition SVM | University of Lubeck | 2008-11-19 | NA | NA |
| Other models | BigChaos 2008 | AT&T | 2008-11-25 | NA | NA |
| Other models | HLBL | University of Toronto | 2008-12-08 | NA | NA |
| Other models | GNN | University of Siena | 2008-12-09 | 1.6e+9 | 1 |
| Other models | BP-DBN | University of Toronto | 2009-01-01 | NA | 1 |
| Other models | RBM Image Classifier | University of Toronto | 2009-04-08 | NA | NA |
| Other models | GPU DBNs | Stanford University | 2009-06-15 | 1.0e+15 | NA |
| Other models | Pragmatic Theory solution (Netflix 2009) | Pragmatic Theory Inc. | 2009-08-01 | NA | NA |
| Other models | BellKor 2008 | AT&T | 2009-08-01 | NA | NA |
| Other models | BigChaos OptiBlend | AT&T | 2009-08-01 | NA | NA |
| Other models | BellKor 2009 | AT&T | 2009-08-01 | NA | NA |
| Other models | MatrixFac for Recommenders | Yahoo Research, AT&T | 2009-08-07 | NA | NA |
| Other models | Two Stage Feature Extraction (MNIST) | New York University (NYU) | 2009-09-01 | 2.1e+13 | NA |
| Other models | BellKor 2007 | AT&T | 2009-09-21 | NA | NA |
| Other models | LCNP MNIST | NA | 2009-11-22 | 4.2e+15 | NA |
| Other models | LCNP LabelMe | University of Bonn | 2009-11-22 | 3.3e+15 | NA |
| Other models | LCNP NORB | NA | 2009-11-22 | 2.5e+15 | NA |
| Other models | Super-vector coding | 2010-01-01 | NA | NA | |
| Other models | Stacked Denoising Autoencoders | 2010-01-03 | NA | NA | |
| Other models | Feedforward NN | 2010-05-13 | 3.5e+14 | NA | |
| Other models | iCCCP | Massachusetts Institute of Technology (MIT) | 2010-06-13 | 1.1e+15 | 1 |
| Other models | ReLU (NORB) | University of Toronto | 2010-06-15 | NA | NA |
| Other models | ReLU (LFW) | University of Toronto | 2010-06-15 | NA | NA |
| Other models | Pooling CNN (Caltech 101) | University of Bonn | 2010-09-15 | 1.2e+15 | NA |
| Other models | Pooling CNN (NORB) | University of Bonn | 2010-09-15 | 1.5e+15 | NA |
| Other models | RNN LM | Johns Hopkins University | 2010-09-26 | 5.4e+16 | NA |
| Other models | YouTube Video Recommendation System | 2010-09-26 | NA | NA | |
| Other models | Deep rectifier networks | 2011-04-13 | NA | NA | |
| Other models | Deep Autoencoders | University of Toronto | 2011-04-29 | 3.7e+16 | 1 |
| Other models | Cross-Lingual POS Tagger | 2011-06-19 | NA | NA | |
| Other models | Vector Space Model | Stanford University | 2011-06-19 | NA | NA |
| Other models | Recursive Neural Network | Stanford University | 2011-06-28 | NA | NA |
| Other models | CNN Committee (MNIST) | IDSIA | 2011-09-18 | 5.2e+16 | 4 |
| Other models | CNN Committee (NIST) | IDSIA | 2011-09-18 | 2.6e+16 | 4 |
| Other models | CNN committee (traffic sign) | IDSIA | 2011-10-03 | 9.9e+14 | 4 |
| Other models | Adaptive Subgrad | 2011-10-03 | NA | NA | |
| Other models | NLP from scratch | NEC Laboratories, Princeton University | 2011-11-08 | NA | NA |
| Other models | Dropout (CIFAR) | University of Toronto | 2012-06-03 | 4.3e+15 | NA |
| Other models | Dropout (TIMIT) | University of Toronto | 2012-06-03 | NA | NA |
| Other models | Dropout (MNIST) | University of Toronto | 2012-06-03 | 6.0e+15 | NA |
| Other models | Dropout (ImageNet) | University of Toronto | 2012-06-03 | 2.7e+17 | NA |
| Other models | Unsupervised High-level Feature Learner | 2012-07-12 | 6.0e+17 | NA | |
| Other models | Context-dependent RNN | 2012-07-27 | NA | NA | |
| Other models | LSTM LM | RWTH Aachen University | 2012-09-09 | 1.7e+16 | NA |
| Other models | AlexNet | University of Toronto | 2012-09-30 | 4.7e+17 | NA |
| Other models | RNN+LDA+KN5+cache | Microsoft, Brno University of Technology | 2012-12-01 | NA | NA |
| Other models | Bayesian automated hyperparameter tuning | 2012-12-02 | NA | NA | |
| Other models | DNN EM segmentation | IDSIA, SUPSI | 2012-12-03 | 4.8e+17 | 4 |
| Other models | DistBelief Vision | 2012-12-03 | NA | NA | |
| Other models | DistBelief Speech | 2012-12-03 | 3.1e+17 | NA | |
| Other models | PreTrans-3L-250H | University of Toronto | 2013-03-22 | NA | NA |
| Other models | Selective Search | University of Trento, University of Amsterdam | 2013-04-02 | NA | NA |
| Other models | Multilingual DNN | 2013-05-26 | NA | NA | |
| Other models | ReLU-Speech | 2013-05-26 | 1.3e+17 | NA | |
| Other models | New York University (NYU) | 2013-08-01 | 2.4e+17 | 1 | |
| Other models | RCTM | University of Oxford | 2013-10-01 | 9.3e+15 | 3 |
| Other models | RNTN | Stanford University | 2013-10-01 | 1.4e+16 | NA |
| Other models | Word2Vec (large) | 2013-10-16 | 3.9e+16 | NA | |
| Other models | Word2Vec (small) | 2013-10-16 | NA | NA | |
| Other models | R-CNN (T-net) | University of California (UC) Berkeley | 2013-11-11 | NA | NA |
| Other models | Visualizing CNNs | New York University (NYU) | 2013-11-12 | 5.3e+17 | NA |
| Other models | DeViSE | 2013-12-05 | NA | NA | |
| Other models | TransE | 2013-12-05 | 1.3e+18 | NA | |
| Other models | RNN for 1B words | 2013-12-11 | NA | 24 | |
| Other models | Network in Network | National University of Singapore | 2013-12-16 | NA | NA |
| Other models | DQN | DeepMind | 2013-12-19 | 2.8e+15 | NA |
| Other models | Image generation | University of Amsterdam | 2013-12-20 | 4.8e+14 | NA |
| Other models | OverFeat | New York University (NYU) | 2013-12-21 | NA | NA |
| Other models | GloVe (32B) | Stanford University | 2014-01-01 | NA | NA |
| Other models | GloVe (6B) | Stanford University | 2014-01-01 | NA | NA |
| Other models | HyperNEAT | University of Texas at Austin | 2014-03-05 | NA | NA |
| Other models | Paragraph Vector | 2014-05-14 | NA | NA | |
| Other models | AdaRNN | Beihang University | 2014-06-01 | NA | NA |
| Other models | Dropout: SVHN | University of Toronto | 2014-06-01 | NA | NA |
| Other models | GRUs | 2014-06-03 | NA | NA | |
| Other models | Two-stream ConvNets for action recognition | University of Oxford | 2014-06-09 | NA | NA |
| Other models | GANs | 2014-06-10 | 5.2e+17 | NA | |
| Other models | SPPNet | 2014-06-18 | 3.4e+18 | NA | |
| Other models | Fragment embedding | Stanford University | 2014-06-21 | NA | NA |
| Other models | RNN-WER | DeepMind, University of Toronto | 2014-06-22 | NA | NA |
| Other models | Multiresolution CNN | Google, Stanford University | 2014-06-23 | NA | NA |
| Other models | DeepFace | Tel Aviv University, Facebook | 2014-06-23 | NA | NA |
| Other models | SmooCT | University College London (UCL) | 2014-07-01 | 6.9e+16 | NA |
| Other models | ACF-WIDER | Chinese Academy of Sciences | 2014-07-15 | 7.6e+13 | 1 |
| Other models | NPD | IEEE | 2014-08-06 | NA | NA |
| Other models | RNNsearch-50* | 2014-09-01 | 1.6e+18 | NA | |
| Other models | VGG19 | University of Oxford | 2014-09-04 | 1.1e+19 | NA |
| Other models | VGG16 | University of Oxford | 2014-09-04 | 1.2e+19 | 4 |
| Other models | Seq2Seq LSTM | 2014-09-10 | 5.6e+19 | NA | |
| Other models | SPN-4+KN5 | 2014-09-14 | 4.4e+16 | 1 | |
| Other models | GoogLeNet / InceptionV1 | 2014-09-17 | 1.5e+18 | NA | |
| Other models | Deeply-supervised nets | Microsoft Research | 2014-09-18 | NA | NA |
| Other models | Spatially-Sparse CNN | University of Warwick | 2014-09-23 | NA | NA |
| Other models | LRCN | 2014-11-07 | NA | NA | |
| Other models | SC-NLM | University of Toronto | 2014-11-10 | NA | NA |
| Other models | Fully Convolutional Networks | University of California (UC) Berkeley | 2014-11-14 | NA | NA |
| Other models | Cascaded LNet-ANet | Chinese University of Hong Kong (CUHK) | 2014-11-28 | NA | NA |
| Other models | TA-CNN | Chinese University of Hong Kong (CUHK) | 2014-11-29 | 1.1e+16 | 1 |
| Other models | SNM-skip | 2014-12-03 | 3.0e+20 | NA | |
| Other models | Fractional Max-Pooling | University of Warwick | 2014-12-18 | 1.0e+17 | NA |
| Other models | ADAM (CIFAR-10) | 2014-12-22 | 6.2e+14 | NA | |
| Other models | VGG-Face | University of Oxford | 2015-01-01 | NA | 4 |
| Other models | MSRA (C, PReLU) | Microsoft Research | 2015-02-06 | 2.4e+19 | NA |
| Other models | TRPO | University of California (UC) Berkeley | 2015-02-19 | NA | NA |
| Other models | DQN-2015 | 2015-02-25 | NA | NA | |
| Other models | genCNN + dyn eval | 2015-03-17 | 3.4e+16 | NA | |
| Other models | TC-DNN-BLSTM-DNN | Carnegie Mellon University (CMU) | 2015-04-06 | 1.9e+17 | 1 |
| Other models | Fast R-CNN | Microsoft Research | 2015-04-30 | NA | NA |
| Other models | U-Net | University of Freiburg | 2015-05-18 | 5.1e+16 | 1 |
| Other models | Faster R-CNN | Microsoft Research | 2015-06-04 | NA | NA |
| Other models | CFSS | 2015-06-07 | NA | 1 | |
| Other models | YOLO | 2015-06-08 | NA | NA | |
| Other models | BatchNorm | 2015-06-15 | NA | NA | |
| Other models | CompACT-Deep | University of California San Diego | 2015-07-19 | NA | 1 |
| Other models | Deep CNN + COTS | IEEE | 2015-07-26 | NA | 1 |
| Other models | DCNN | University of Maryland, Rutgers University | 2015-08-07 | 4.8e+17 | 1 |
| Other models | BPE | University of Edinburgh | 2015-08-31 | NA | NA |
| Other models | Deep Deterministic Policy Gradients | Google DeepMind | 2015-09-09 | NA | NA |
| Other models | AlphaGo Fan | DeepMind | 2015-10-01 | 3.8e+20 | NA |
| Other models | SAF R-CNN | 2015-10-28 | 1.2e+19 | 1 | |
| Other models | 3DDFA | 2015-11-23 | NA | NA | |
| Other models | Multi-scale Dilated CNN | Princeton University, Intel Labs | 2015-11-23 | NA | NA |
| Other models | Netflix Recommender System | Netflix | 2015-12-01 | NA | NA |
| Other models | Inception v3 | Google, University College London (UCL) | 2015-12-02 | 1.0e+20 | NA |
| Other models | SSD | NA | 2015-12-08 | NA | NA |
| Other models | ResNet-110 (CIFAR-10) | Microsoft | 2015-12-10 | NA | NA |
| Other models | ResNet-152 (ImageNet) | Microsoft | 2015-12-10 | 1.0e+19 | NA |
| Other models | ResNet-101 (ImageNet) | Microsoft | 2015-12-10 | 7.0e+18 | NA |
| Other models | Advantage Learning | Google DeepMind | 2015-12-15 | NA | NA |
| Other models | Variational (untied weights, MC) LSTM (Large) | University of Cambridge | 2015-12-16 | 5.9e+15 | NA |
| Other models | AlphaGo Lee | DeepMind | 2016-01-27 | 1.9e+21 | NA |
| Other models | A3C FF hs | 2016-02-04 | NA | NA | |
| Other models | Inception-ResNet-V2 | 2016-02-23 | NA | NA | |
| Other models | Inceptionv4 | 2016-02-23 | NA | NA | |
| Other models | SqueezeNet | 2016-02-24 | NA | NA | |
| Other models | Double DQN | Google DeepMind | 2016-03-02 | NA | NA |
| Other models | Named Entity Recognition model | Carnegie Mellon University (CMU) | 2016-03-04 | 9.7e+16 | 1 |
| Other models | Template Adaptation | University of Oxford | 2016-03-12 | NA | NA |
| Other models | Dueling DQN | Google DeepMind | 2016-04-05 | NA | NA |
| Other models | Gated HORNN (3rd order) | York University | 2016-04-30 | NA | NA |
| Other models | LRR-4X | UC Irvine | 2016-05-08 | NA | 1 |
| Other models | PixelCNN | Google DeepMind | 2016-06-16 | NA | 32 |
| Other models | CMS-RCNN | IEEE | 2016-06-17 | NA | NA |
| Other models | Segmental RNN | 2016-06-20 | NA | NA | |
| Other models | R-FCN | Tsinghua University, Microsoft Research | 2016-06-21 | 7.2e+17 | NA |
| Other models | CCL | 2016-06-27 | NA | NA | |
| Other models | Character-enriched word2vec | Facebook AI Research | 2016-07-15 | NA | NA |
| Other models | SimpleNet | 2016-08-22 | NA | NA | |
| Other models | DenseNet-264 | 2016-08-25 | NA | NA | |
| Other models | LF-MMI | Johns Hopkins University, Cornell University | 2016-09-08 | NA | NA |
| Other models | MS-ensemble-speech-recognition | Microsoft | 2016-09-12 | NA | NA |
| Other models | WaveNet | Google DeepMind | 2016-09-12 | NA | NA |
| Other models | Youtube recommendation model | 2016-09-15 | NA | NA | |
| Other models | ResNet-200 | Microsoft Research Asia | 2016-09-17 | 3.0e+19 | NA |
| Other models | ResNet-1001 | Microsoft | 2016-09-17 | NA | NA |
| Other models | Wide Residual Network | Université Paris-Est | 2016-09-19 | NA | NA |
| Other models | GNMT | 2016-09-26 | 6.6e+21 | NA | |
| Other models | Pointer Sentinel-LSTM (medium) | MetaMind Inc, Salesforce | 2016-09-26 | 7.5e+15 | NA |
| Other models | Xception | 2016-10-07 | 4.4e+20 | 60 | |
| Other models | GAWWN | 2016-10-08 | NA | NA | |
| Other models | SPIDER2 | 2016-10-28 | 1.8e+16 | NA | |
| Other models | VD-LSTM+REAL Large | Salesforce Research, Stanford University | 2016-11-04 | 2.1e+16 | NA |
| Other models | BIDAF | 2016-11-05 | 3.5e+18 | 8 | |
| Other models | NAS with base 8 and shared embeddings | Google Brain | 2016-11-05 | 1.0e+16 | NA |
| Other models | DLDL (PASCAL) | University of Oxford | 2016-11-06 | NA | 1 |
| Other models | DTN (Domain Transfer Network) | Facebook AI Research | 2016-11-07 | NA | NA |
| Other models | DAC-CSR | Jiangnan University, University of Surrey | 2016-11-16 | NA | NA |
| Other models | ResNeXt-50 | University of California San Diego, Facebook | 2016-11-16 | NA | NA |
| Other models | PolyNet | Chinese University of Hong Kong (CUHK) | 2016-11-17 | 6.4e+19 | 32 |
| Other models | Image-to-image cGAN | University of California (UC) Berkeley | 2016-11-21 | NA | NA |
| Other models | PointNet | Stanford University | 2016-12-02 | NA | NA |
| Other models | Elastic weight consolidation | DeepMind | 2016-12-02 | NA | NA |
| Other models | GAN-Advancer | OpenAI | 2016-12-05 | NA | NA |
| Other models | HR-ResNet101 | Carnegie Mellon University (CMU) | 2016-12-13 | 7.1e+18 | NA |
| Other models | 3DMM-CNN | University of Southern California | 2016-12-15 | NA | 1 |
| Other models | EnhanceNet | Max Planck Institute for Intelligent Systems | 2016-12-23 | 1.3e+17 | 1 |
| Other models | YOLOv2 | 2016-12-25 | NA | NA | |
| Other models | DeepStack | 2017-01-06 | 1.4e+19 | 20 | |
| Other models | OR-WideResNet | 2017-01-07 | NA | NA | |
| Other models | MoE-Multi | Jagiellonian University, Google Brain | 2017-01-23 | 9.4e+19 | 64 |
| Other models | DnCNN | 2017-02-01 | NA | NA | |
| Other models | Prototypical networks | University of Toronto, Twitter | 2017-03-15 | NA | NA |
| Other models | Mask R-CNN | Facebook AI Research | 2017-03-30 | NA | NA |
| Other models | WGAN-GP | 2017-03-31 | NA | NA | |
| Other models | MobileNet | 2017-04-17 | NA | NA | |
| Other models | DeepLab (2017) | 2017-04-27 | NA | NA | |
| Other models | Mnemonic Reader | Fudan University, Microsoft Research | 2017-05-08 | NA | NA |
| Other models | SRGAN | 2017-05-25 | NA | NA | |
| Other models | Inflated 3D ConvNet | DeepMind, University of Oxford | 2017-06-01 | NA | NA |
| Other models | PointNet++ | Stanford University | 2017-06-07 | NA | NA |
| Other models | Reading Twice for NLU | DeepMind | 2017-06-08 | NA | NA |
| Other models | EDSR | Seoul National University | 2017-06-10 | NA | NA |
| Other models | Transformer | Google Research, Google Brain | 2017-06-12 | 7.4e+18 | 8 |
| Other models | HRA | Maluuba, Microsoft | 2017-06-13 | NA | NA |
| Other models | DeepLabV3 | 2017-06-17 | NA | NA | |
| Other models | NoisyNet-Dueling | DeepMind | 2017-06-30 | NA | NA |
| Other models | ShuffleNet v1 | Megvii Inc | 2017-07-03 | NA | NA |
| Other models | DeepLoc | 2017-07-07 | 5.8e+17 | 1 | |
| Other models | JFT | 2017-07-10 | 8.4e+20 | 50 | |
| Other models | AWD-LSTM | DeepMind, University of Oxford | 2017-07-18 | NA | NA |
| Other models | PSPNet | Chinese University of Hong Kong (CUHK) | 2017-07-21 | NA | NA |
| Other models | NASNet-A | Google Brain | 2017-07-21 | NA | NA |
| Other models | ConvS2S (ensemble of 8 models) | Meta AI | 2017-07-25 | 5.6e+19 | NA |
| Other models | GSM | Peking University, Microsoft Research | 2017-07-30 | NA | NA |
| Other models | RetinaNet-R101 | Facebook AI Research | 2017-08-07 | 2.1e+18 | 8 |
| Other models | Salesforce Research | 2017-08-07 | 3.0e+17 | NA | |
| Other models | RetinaNet-R50 | Facebook AI Research | 2017-08-07 | NA | NA |
| Other models | OpenAI TI7 DOTA 1v1 | OpenAI | 2017-08-11 | 6.0e+20 | NA |
| Other models | EI-REHN-1000D | 2017-08-14 | 1.1e+16 | NA | |
| Other models | NeuMF (Pinterest) | 2017-08-16 | NA | NA | |
| Other models | Libratus | Carnegie Mellon University (CMU) | 2017-08-19 | 5.5e+20 | NA |
| Other models | Ben-Gurion University of the Negev | 2017-08-29 | 4.6e+17 | NA | |
| Other models | SENet (ImageNet) | 2017-09-05 | NA | NA | |
| Other models | PyramidNet | 2017-09-06 | 2.3e+15 | NA | |
| Other models | ISS | Duke University, Microsoft | 2017-09-15 | 3.4e+15 | NA |
| Other models | LSTM + dynamic eval | University of Edinburgh | 2017-09-21 | NA | NA |
| Other models | AWD-LSTM+WT+Cache+IOG (WT2) | NTT Communication Science Laboratories | 2017-09-26 | 3.2e+15 | NA |
| Other models | Rainbow DQN | DeepMind | 2017-10-06 | NA | 1 |
| Other models | AlphaGo Zero | DeepMind | 2017-10-18 | 6.5e+20 | NA |
| Other models | AlphaGo Master | DeepMind | 2017-10-19 | 3.4e+20 | NA |
| Other models | ProgressiveGAN | NVIDIA | 2017-10-27 | NA | NA |
| Other models | PhraseCond | 2017-10-28 | NA | NA | |
| Other models | S-Norm | 2017-10-29 | NA | NA | |
| Other models | Fraternal dropout + AWD-LSTM 3-layer (WT2) | 2017-10-31 | 3.1e+17 | NA | |
| Other models | DCN+ | Salesforce Research | 2017-10-31 | NA | NA |
| Other models | VQ-VAE | DeepMind | 2017-11-02 | NA | NA |
| Other models | AWD-LSTM-MoS + dynamic evaluation (WT2, 2017) | Carnegie Mellon University (CMU) | 2017-11-10 | 3.4e+18 | NA |
| Other models | TriNet | 2017-11-21 | NA | NA | |
| Other models | DL scaling LM | Baidu | 2017-12-01 | NA | NA |
| Other models | DL scaling speech | Baidu | 2017-12-01 | NA | NA |
| Other models | DL scaling Image | Baidu | 2017-12-01 | NA | NA |
| Other models | ELMo | 2018-02-01 | 3.3e+15 | NA | |
| Other models | QRNN | Salesforce Research | 2018-02-01 | 6.9e+17 | NA |
| Other models | IMPALA | DeepMind | 2018-02-05 | 1.7e+20 | 1 |
| Other models | DeepLabV3+ | 2018-02-07 | NA | NA | |
| Other models | TCN (P-MNIST) | Carnegie Mellon University (CMU), Intel Labs | 2018-02-15 | NA | NA |
| Other models | Chinese - English translation | Microsoft | 2018-03-01 | NA | NA |
| Other models | 4 layer QRNN (h=2500) | Salesforce Research | 2018-03-22 | 5.9e+17 | 1 |
| Other models | YOLOv3 | University of Washington | 2018-04-08 | 1.3e+19 | NA |
| Other models | ResNeXt-101 32x48d | 2018-05-02 | 8.7e+21 | 336 | |
| Other models | Dropout-LSTM+Noise(Bernoulli) (WT2) | 2018-05-03 | 1.3e+17 | NA | |
| Other models | aLSTM(depth-2)+RecurrentPolicy (WT2) | 2018-05-22 | 7.3e+16 | NA | |
| Other models | GPT-1 | OpenAI | 2018-06-01 | 1.8e+19 | 8 |
| Other models | Relational Memory Core | DeepMind, University College London (UCL) | 2018-06-05 | NA | NA |
| Other models | MobileNetV2 | 2018-06-18 | NA | NA | |
| Other models | FTW (For The Win) | DeepMind | 2018-07-03 | 3.5e+19 | NA |
| Other models | Big-Little Net (speech) | IBM | 2018-07-10 | 4.3e+17 | NA |
| Other models | Big-Little Net | IBM | 2018-07-10 | 2.5e+17 | NA |
| Other models | AWD-LSTM-MoS+PDR + dynamic evaluation (WT2) | IBM | 2018-08-14 | NA | NA |
| Other models | Big Transformer for Back-Translation | Facebook AI Research, Google Brain | 2018-08-28 | 4.8e+20 | 128 |
| Other models | (ensemble): AWD-LSTM-DOC (fin) × 5 (WT2) | 2018-08-30 | 6.7e+17 | NA | |
| Other models | NetSurfP-2.0 | 2018-09-10 | NA | NA | |
| Other models | Transformer + Simple Recurrent Unit | 2018-09-17 | 1.1e+19 | 8 | |
| Other models | AWD-LSTM-MoS + dynamic evaluation (WT2, 2018) | Peking University, Microsoft Research Asia | 2018-09-18 | NA | NA |
| Other models | LSTM+NeuralCache | KU Leuven, ESAT - PSI, Apple | 2018-09-24 | 9.8e+14 | NA |
| Other models | Transformer (Adaptive Input Embeddings) WT103 | Facebook AI Research | 2018-09-28 | 4.5e+19 | 8 |
| Other models | MetaMimic | 2018-10-11 | NA | NA | |
| Other models | BERT-Large | 2018-10-11 | 2.8e+20 | 64 | |
| Other models | TrellisNet | 2018-10-15 | 2.8e+18 | NA | |
| Other models | MemoReader | Samsung, Korea University | 2018-10-31 | NA | NA |
| Other models | Google Brain | 2018-11-05 | 6.8e+19 | 64 | |
| Other models | Mesh-TensorFlow Transformer 4.9B (language) | Google Brain | 2018-11-05 | 1.6e+20 | 256 |
| Other models | Fine-tuned-AWD-LSTM-DOC (fin) | Samsung R&D Institute Russia | 2018-11-12 | 5.2e+16 | NA |
| Other models | Multi-cell LSTM | University of Hyderabad | 2018-11-15 | 2.0e+15 | NA |
| Other models | GPipe (Transformer) | 2018-11-16 | NA | NA | |
| Other models | SPN (ImageNet 128) | Google Brain, DeepMind | 2018-12-04 | NA | NA |
| Other models | StyleGAN | NVIDIA | 2018-12-12 | 3.9e+16 | 8 |
| Other models | Transformer ELMo | 2019-01-01 | NA | NA | |
| Other models | Transformer-XL (257M) | 2019-01-09 | 3.8e+20 | 32 | |
| Other models | MT-DNN | Microsoft | 2019-01-31 | NA | NA |
| Other models | Hanabi 4 player | 2019-02-01 | 4.3e+18 | NA | |
| Other models | GPT-2 (1.5B) | OpenAI | 2019-02-14 | 1.9e+21 | NA |
| Other models | KataGo | Jane Street | 2019-02-27 | 2.3e+19 | NA |
| Other models | NMT Transformer 437M | Google, Bar-Ilan University | 2019-02-28 | NA | NA |
| Other models | SciBERT | Allen Institute for AI | 2019-03-26 | 8.9e+19 | 4 |
| Other models | Cross-lingual alignment | 2019-04-04 | 2.6e+18 | NA | |
| Other models | True-Regularization+Finetune+Dynamic-Eval | Mobvoi, Williams College | 2019-04-08 | NA | NA |
| Other models | WeNet (Penn Treebank) | Amazon | 2019-04-08 | 7.3e+17 | 1 |
| Other models | Transformer-XL + RMS dynamic eval | University of Edinburgh | 2019-04-17 | NA | NA |
| Other models | BERT-Large-CAS (PTB+WT2+WT103) | Amazon | 2019-04-20 | 1.5e+20 | NA |
| Other models | MuseNet | OpenAI | 2019-04-25 | 2.2e+20 | NA |
| Other models | Neuro-Symbolic Concept Learner | 2019-04-26 | NA | NA | |
| Other models | RaptorX-Contact | Toyota Technological Institute at Chicago | 2019-05-02 | NA | NA |
| Other models | ResNeXt-101 Billion-scale | Facebook AI | 2019-05-02 | NA | NA |
| Other models | AWD-LSTM-DRILL + dynamic evaluation† (WT2) | IDIAP | 2019-05-14 | 4.1e+17 | NA |
| Other models | CPC v2 | 2019-05-22 | NA | NA | |
| Other models | EfficientNet-L2 | 2019-05-28 | NA | NA | |
| Other models | DLRM-2020 | Facebook AI | 2019-05-31 | 4.0e+18 | NA |
| Other models | XLNet | 2019-06-01 | 6.2e+21 | NA | |
| Other models | XLM | 2019-06-01 | NA | NA | |
| Other models | Transformer-XL Large + Phrase Induction | 2019-06-04 | 3.8e+20 | NA | |
| Other models | AWD-LSTM + MoS + Partial Shuffled | University of Texas at Austin | 2019-06-10 | 3.2e+17 | NA |
| Other models | Char-CNN-BiLSTM | Capital One | 2019-06-13 | NA | NA |
| Other models | FixRes ResNeXt-101 WSL | Facebook AI | 2019-06-14 | NA | NA |
| Other models | PG-SWGAN | ETH Zurich | 2019-06-15 | NA | NA |
| Other models | LaNet-L (CIFAR-10) | Brown University, Facebook | 2019-06-17 | NA | NA |
| Other models | Walking Minotaur robot | 2019-06-19 | NA | NA | |
| Other models | BigBiGAN | 2019-07-04 | NA | NA | |
| Other models | Pluribus | Facebook AI Research | 2019-07-11 | 6.6e+16 | NA |
| Other models | EN^2AS with performance reward | 2019-07-22 | NA | NA | |
| Other models | trRosetta | 2019-08-22 | 3.8e+19 | 1 | |
| Other models | UDSMProt | Fraunhofer Heinrich Hertz Institute | 2019-09-04 | 6.4e+17 | NA |
| Other models | Mogrifier (d2, MoS2, MC) + dynamic eval | DeepMind, University of Oxford | 2019-09-04 | NA | NA |
| Other models | Megatron-LM (1.2B) | NVIDIA | 2019-09-17 | 1.1e+22 | 1 |
| Other models | Megatron-LM (8.3B) | NVIDIA | 2019-09-17 | 9.1e+21 | 512 |
| Other models | Megatron-BERT | NVIDIA | 2019-09-17 | 2.2e+22 | 512 |
| Other models | Adaptive Inputs + LayerDrop | Facebook AI Research, LORIA | 2019-09-25 | NA | NA |
| Other models | ALBERT | 2019-09-26 | NA | NA | |
| Other models | AlphaX-1 | Facebook AI Research, Brown University | 2019-10-02 | 8.9e+17 | NA |
| Other models | DistilBERT | Hugging Face | 2019-10-02 | 1.2e+19 | NA |
| Other models | M4-50B | 2019-10-11 | NA | NA | |
| Other models | T5-11B | 2019-10-23 | 3.3e+22 | 512 | |
| Other models | T5-3B | 2019-10-23 | 9.0e+21 | NA | |
| Other models | BART-large | Facebook AI | 2019-10-29 | NA | NA |
| Other models | AlphaStar | DeepMind | 2019-10-30 | 1.1e+23 | 384 |
| Other models | Base LM + kNN LM + Continuous Cache | Stanford University, Facebook AI Research | 2019-11-01 | 3.1e+19 | NA |
| Other models | XLM-RoBERTa | Facebook AI | 2019-11-05 | 2.1e+22 | 500 |
| Other models | CamemBERT | Facebook, INRIA, Sorbonne University | 2019-11-10 | 8.3e+20 | NA |
| Other models | Sandwich Transformer | Allen Institute for AI, Facebook AI Research | 2019-11-10 | 2.4e+19 | NA |
| Other models | Noisy Student (L2) | Carnegie Mellon University (CMU), Google | 2019-11-11 | 2.6e+22 | 1,024 |
| Other models | MoCo | Facebook AI | 2019-11-13 | NA | NA |
| Other models | MuZero | DeepMind | 2019-11-19 | 4.8e+19 | NA |
| Other models | Transformer - LibriVox + Decoding/Rescoring | 2019-11-19 | NA | NA | |
| Other models | Photo-Geometric Autoencoder | University of Oxford | 2019-11-25 | NA | NA |
| Other models | Transformer-XL DeFINE (141M) | 2019-11-27 | 1.7e+18 | NA | |
| Other models | StyleGAN2 | NVIDIA, Aalto University | 2019-12-03 | NA | 8 |
| Other models | StarGAN v2 | 2019-12-04 | NA | NA | |
| Other models | MMLSTM (PTB) | 2019-12-05 | 5.8e+16 | NA | |
| Other models | MMLSTM (WT-2) | 2019-12-05 | 1.9e+17 | NA | |
| Other models | OpenAI Five Rerun | OpenAI | 2019-12-13 | 1.3e+22 | 512 |
| Other models | DD-PPO | 2019-12-19 | 7.8e+20 | 64 | |
| Other models | Big Transfer (BiT-L) | Google Brain | 2019-12-24 | NA | NA |
| Other models | AlphaFold | DeepMind | 2020-01-15 | 1.0e+20 | NA |
| Other models | ContextNet + Noisy Student | 2020-01-19 | 8.2e+21 | NA | |
| Other models | Meena | Google Brain | 2020-01-28 | 1.1e+23 | 1,024 |
| Other models | Theseus 6/768 | 2020-02-07 | NA | NA | |
| Other models | Perceiver IO (optical flow) | DeepMind | 2020-02-08 | NA | NA |
| Other models | TaLK Convolution | Carleton University | 2020-02-08 | 2.7e+19 | 8 |
| Other models | ALBERT-xxlarge | 2020-02-09 | 2.4e+21 | 512 | |
| Other models | SimCLR | Google Brain | 2020-02-13 | NA | NA |
| Other models | Turing-NLG | Microsoft | 2020-02-13 | 1.6e+22 | 256 |
| Other models | Feedback Transformer | 2020-02-21 | 7.7e+18 | NA | |
| Other models | TCAN (WT2) | Nanjing University, Ant Group | 2020-02-28 | NA | NA |
| Other models | TransformerXL + spectrum control | 2020-03-11 | 2.6e+19 | 4 | |
| Other models | Routing Transformer (WT-103) | Google Research | 2020-03-12 | NA | NA |
| Other models | Tensor-Transformer(1core)+PN (WT103) | University of California (UC) Berkeley | 2020-03-17 | 1.6e+18 | NA |
| Other models | ELECTRA | Stanford University, Google, Google Brain | 2020-03-23 | 3.1e+21 | NA |
| Other models | MetNet | 2020-03-24 | 9.5e+18 | 256 | |
| Other models | Agent57 | DeepMind | 2020-03-30 | NA | NA |
| Other models | CURL | University of California (UC) Berkeley | 2020-04-08 | NA | NA |
| Other models | Go-explore | Uber AI, OpenAI | 2020-04-27 | NA | NA |
| Other models | Once for All | 2020-04-29 | 6.2e+20 | NA | |
| Other models | UnifiedQA | 2020-05-02 | 1.6e+19 | 8 | |
| Other models | NAS+ESS (23M) | 2020-05-06 | NA | 1 | |
| Other models | ContextNet | 2020-05-07 | NA | NA | |
| Other models | Conformer | 2020-05-16 | NA | NA | |
| Other models | Retrieval-Augmented Generator | 2020-05-22 | NA | NA | |
| Other models | DETR | 2020-05-26 | 4.0e+20 | NA | |
| Other models | GShard (dense) | 2020-06-30 | 4.8e+22 | 1,024 | |
| Other models | SemExp | 2020-07-02 | NA | NA | |
| Other models | Hopfield Networks (2020) | 2020-07-16 | NA | NA | |
| Other models | EfficientDet | Google Brain | 2020-07-27 | NA | NA |
| Other models | DeLighT | 2020-08-03 | 3.8e+18 | 8 | |
| Other models | ERNIE-GEN (large) | Baidu | 2020-08-06 | 2.0e+20 | NA |
| Other models | ProBERTa | 2020-09-01 | 9.7e+18 | 4 | |
| Other models | LUKE | 2020-10-02 | 1.8e+22 | 16 | |
| Other models | Conformer + Wav2vec 2.0 + Noisy Student | Google, Google Research, Google Brain | 2020-10-20 | 7.6e+21 | 256 |
| Other models | mT5-XXL | Google, Google Research | 2020-10-20 | 8.2e+22 | NA |
| Other models | German ELECTRA Large | deepset, Bayerische Staatsbibliothek Muenchen | 2020-10-21 | 1.4e+21 | 64 |
| Other models | ViT-Huge/14 | Google Brain, Google Research | 2020-10-22 | 4.3e+21 | NA |
| Other models | wave2vec 2.0 LARGE | 2020-10-22 | 3.9e+21 | NA | |
| Other models | ViT-Base/32 | Google Brain | 2020-10-22 | NA | NA |
| Other models | KEPLER | 2020-11-23 | 1.7e+21 | NA | |
| Other models | AlphaFold 2 | DeepMind | 2020-11-30 | 3.0e+21 | NA |
| Other models | CPM-Large | 2020-12-01 | 2.6e+20 | 64 | |
| Other models | ESM1b | 2020-12-15 | 5.1e+21 | 128 | |
| Other models | VQGAN + CLIP | Heidelberg University | 2020-12-17 | NA | NA |
| Other models | DensePhrases | Korea University, Princeton University | 2020-12-23 | 2.1e+18 | 8 |
| Other models | CT-MoS (WT2) | Google, National Tsing Hua University | 2020-12-25 | 5.4e+17 | 4 |
| Other models | ERNIE-Doc (247M) | Baidu | 2020-12-31 | 3.0e+19 | NA |
| Other models | CLIP (ViT L/14@336px) | OpenAI | 2021-01-05 | 1.0e+22 | 256 |
| Other models | CLIP (ResNet-50) | OpenAI | 2021-01-05 | NA | NA |
| Other models | DALL-E | OpenAI | 2021-01-05 | 4.7e+22 | 1,024 |
| Other models | BigSSL | Google, Apple | 2021-01-10 | NA | NA |
| Other models | Switch | 2021-01-11 | 8.2e+22 | 1,024 | |
| Other models | DeiT-B | Meta AI, Sorbonne University | 2021-01-15 | 7.9e+19 | NA |
| Other models | top-down frozen classifier | 2021-02-09 | NA | NA | |
| Other models | DLWP | University of Washington, Microsoft Research | 2021-02-09 | 5.7e+18 | 1 |
| Other models | MSA Transformer | 2021-02-13 | 5.5e+21 | 32 | |
| Other models | Rational DQN Average | TU Darmstadt | 2021-02-18 | NA | NA |
| Other models | SRU++ Large | ASAPP | 2021-02-24 | 2.1e+19 | NA |
| Other models | Meta Pseudo Labels | Google Brain, Google AI | 2021-03-01 | 4.8e+22 | 1,024 |
| Other models | M6-T | Alibaba | 2021-03-05 | 5.5e+21 | 480 |
| Other models | Generative BST | Facebook AI Research | 2021-03-05 | 1.4e+22 | NA |
| Other models | Unicorn | Allen Institute for AI | 2021-03-24 | NA | NA |
| Other models | PLUG | Alibaba | 2021-04-19 | 3.6e+22 | 128 |
| Other models | ProtBERT-BFD | 2021-05-04 | 3.9e+22 | 1,024 | |
| Other models | ProtT5-XL-U50 | 2021-05-04 | 1.9e+22 | 256 | |
| Other models | ADM | OpenAI | 2021-05-11 | 6.2e+21 | NA |
| Other models | MedBERT | 2021-05-20 | 9.5e+18 | 1 | |
| Other models | Transformer local-attention (NesT-B) | Google Cloud, Google Research | 2021-05-26 | 2.4e+19 | NA |
| Other models | CogView | Tsinghua University, Alibaba DAMO Academy | 2021-05-26 | 2.7e+22 | 512 |
| Other models | ByT5-XXL | Google, Google Research | 2021-05-28 | 8.1e+22 | NA |
| Other models | CoAtNet | Google, Google Research, Google Brain | 2021-06-09 | 4.3e+22 | NA |
| Other models | EMDR | 2021-06-09 | 1.9e+21 | NA | |
| Other models | DeBERTa | Microsoft | 2021-06-10 | 2.6e+22 | 256 |
| Other models | University of California (UC) Berkeley | 2021-06-11 | 7.8e+19 | NA | |
| Other models | ALIGN | Google Research | 2021-06-11 | 2.6e+22 | 512 |
| Other models | StyleGAN3-R | NVIDIA, Aalto University | 2021-06-21 | 2.4e+21 | 8 |
| Other models | StyleGAN3-T | NVIDIA, Aalto University | 2021-06-21 | 1.7e+21 | 8 |
| Other models | EfficientNetV2-XL | Google, Google Brain | 2021-06-23 | 9.6e+19 | 16 |
| Other models | Fold2Seq | IBM, Texas A&M | 2021-06-24 | 1.4e+17 | 2 |
| Other models | Adaptive Input Transformer + RD | Microsoft Research Asia, Soochow University | 2021-06-28 | 8.6e+19 | 8 |
| Other models | ERNIE 3.0 | Baidu | 2021-07-05 | 2.2e+22 | 384 |
| Other models | Codex | OpenAI | 2021-07-07 | 7.3e+22 | NA |
| Other models | HuBERT | Facebook AI Research | 2021-07-27 | 5.5e+21 | NA |
| Other models | GOAT | DeepMind | 2021-07-27 | 2.4e+22 | NA |
| Other models | SEER | Facebook AI Research, INRIA | 2021-07-29 | 1.8e+22 | 512 |
| Other models | 6-Act Tether | 2021-08-03 | NA | NA | |
| Other models | YOLOX-X | Megvii Inc | 2021-08-06 | 6.3e+20 | 8 |
| Other models | W2v-BERT | 2021-08-07 | NA | NA | |
| Other models | Zidong Taichu | 2021-08-11 | 8.0e+20 | NA | |
| Other models | Jurassic-1-Jumbo | AI21 Labs | 2021-08-11 | 3.7e+23 | NA |
| Other models | DNABERT | Northeastern University | 2021-08-15 | 1.1e+20 | NA |
| Other models | XLMR-XXL | Facebook AI Research | 2021-08-17 | 3.4e+22 | NA |
| Other models | FLAN 137B | Google Research | 2021-09-03 | 2.0e+24 | NA |
| Other models | MEB | Microsoft | 2021-09-04 | NA | NA |
| Other models | PermuteFormer | Peking University | 2021-09-06 | 2.8e+18 | 8 |
| Other models | HyperCLOVA 204B | NAVER | 2021-09-10 | 2.0e+23 | NA |
| Other models | PLATO-XL | Baidu | 2021-09-20 | 9.9e+21 | 256 |
| Other models | TrOCR | Beihang University, Microsoft Research Asia | 2021-09-21 | NA | 32 |
| Other models | Turing ULRv5 | Microsoft | 2021-09-28 | 2.9e+22 | 256 |
| Other models | AlphaFold-Multimer | Google DeepMind, DeepMind | 2021-10-04 | 4.4e+21 | 64 |
| Other models | Yuan 1.0 | Inspur | 2021-10-12 | 3.5e+23 | 2,128 |
| Other models | base LM+GNN+kNN | 2021-10-17 | 5.3e+19 | NA | |
| Other models | Eve | Harvard Medical School, University of Oxford | 2021-10-27 | NA | NA |
| Other models | EfficientZero | 2021-10-30 | NA | NA | |
| Other models | S4 | Stanford University | 2021-10-31 | 7.8e+19 | 8 |
| Other models | Projected GAN | Heidelberg University | 2021-11-01 | 1.0e+19 | NA |
| Other models | CodeT5-base | Salesforce, Nanyang Technological University | 2021-11-01 | 1.6e+21 | NA |
| Other models | Masked Autoencoders ViT-H | Facebook AI Research | 2021-11-11 | 4.6e+20 | NA |
| Other models | ViT-G/14 (LiT) | Google Research | 2021-11-15 | NA | NA |
| Other models | Swin Transformer V2 (SwinV2-G) | Microsoft Research Asia | 2021-11-18 | 1.1e+21 | NA |
| Other models | BASIC-L | 2021-11-19 | 4.1e+22 | NA | |
| Other models | Florence | Microsoft | 2021-11-22 | 4.8e+22 | 512 |
| Other models | NÜWA | Microsoft Research, Peking University | 2021-11-24 | 7.2e+21 | NA |
| Other models | T-NLRv5 XXL | Microsoft | 2021-12-03 | NA | NA |
| Other models | Student of Games | DeepMind | 2021-12-06 | 3.7e+22 | NA |
| Other models | Gopher (280B) | DeepMind | 2021-12-08 | 6.3e+23 | 4,096 |
| Other models | GLaM | 2021-12-13 | 3.6e+23 | 1,024 | |
| Other models | LongT5 | Google Research | 2021-12-15 | NA | 128 |
| Other models | Contriever | 2021-12-16 | 1.6e+20 | NA | |
| Other models | XGLM-7.5B | Meta AI, Facebook AI Research | 2021-12-20 | 2.2e+22 | 256 |
| Other models | LDM-1.45B | Heidelberg University, Runway | 2021-12-20 | NA | NA |
| Other models | ERNIE 3.0 Titan | Baidu, Peng Cheng Laboratory | 2021-12-23 | 1.0e+24 | 1,920 |
| Other models | ERNIE-ViLG | Baidu | 2021-12-31 | NA | NA |
| Other models | Detic | Meta AI, University of Texas at Austin | 2022-01-07 | 2.3e+19 | 32 |
| Other models | data2vec (vision) | Meta AI | 2022-01-20 | NA | NA |
| Other models | data2vec (language) | Meta AI | 2022-01-20 | NA | NA |
| Other models | data2vec (speech) | Meta AI | 2022-01-20 | NA | NA |
| Other models | AbLang (heavy sequences) | University of Oxford | 2022-01-22 | NA | NA |
| Other models | OntoProtein | Zhejiang University (ZJU) | 2022-01-23 | NA | NA |
| Other models | InstructGPT 175B | OpenAI | 2022-01-27 | 3.2e+23 | NA |
| Other models | InstructGPT 1.3B | OpenAI | 2022-01-27 | NA | NA |
| Other models | InstructGPT 6B | OpenAI | 2022-01-27 | NA | NA |
| Other models | AlphaCode | DeepMind | 2022-02-02 | 2.4e+23 | 3,750 |
| Other models | RETRO-7B | DeepMind | 2022-02-07 | 1.7e+22 | NA |
| Other models | GPT-NeoX-20B | EleutherAI | 2022-02-09 | 9.3e+22 | 96 |
| Other models | ProteinBERT | 2022-02-10 | 6.5e+19 | 1 | |
| Other models | LaMDA | 2022-02-10 | 3.6e+23 | 1,024 | |
| Other models | Midjourney V1 | Midjourney | 2022-02-15 | NA | NA |
| Other models | ST-MoE | Google, Google Brain, Google Research | 2022-02-17 | 2.9e+23 | NA |
| Other models | FourCastNet | 2022-02-22 | 3.5e+20 | 64 | |
| Other models | PolyCoder | Carnegie Mellon University (CMU) | 2022-02-26 | 1.1e+21 | NA |
| Other models | DeepNet | Microsoft Research | 2022-03-01 | NA | NA |
| Other models | Statement Curriculum Learning | OpenAI | 2022-03-02 | NA | NA |
| Other models | MegaSyn | Collaborations Pharmaceuticals | 2022-03-07 | NA | NA |
| Other models | ViT-G (model soup) | 2022-03-10 | 3.4e+21 | NA | |
| Other models | Segatron-XL large, M=384 + HCP | Microsoft Research, University of Waterloo | 2022-03-21 | 2.6e+19 | NA |
| Other models | Make-A-Scene | Meta AI | 2022-03-24 | 6.4e+21 | NA |
| Other models | Chinchilla | DeepMind | 2022-03-29 | 5.8e+23 | NA |
| Other models | DALL·E 2 | OpenAI | 2022-04-06 | 3.4e+23 | NA |
| Other models | BERT-RBP | Waseda University | 2022-04-07 | 1.4e+20 | NA |
| Other models | Stable Diffusion (LDM-KL-8-G) | 2022-04-13 | 5.0e+22 | 256 | |
| Other models | Sparse all-MLP | Meta AI | 2022-04-14 | 5.3e+20 | NA |
| Other models | XMC-GAN | Google Research | 2022-04-14 | NA | NA |
| Other models | Flamingo | DeepMind | 2022-04-29 | 2.2e+23 | 1,536 |
| Other models | OPT-175B | Meta AI | 2022-05-02 | 4.3e+23 | 1,024 |
| Other models | DeBERTaV3large + KEAR | Microsoft | 2022-05-04 | NA | NA |
| Other models | UL2 | Google Research, Google Brain | 2022-05-10 | 1.2e+23 | 512 |
| Other models | Gato | DeepMind | 2022-05-12 | 4.0e+21 | 256 |
| Other models | SimCSE | Princeton University, Tsinghua University | 2022-05-18 | NA | NA |
| Other models | Imagen | Google Brain | 2022-05-23 | 1.5e+22 | 256 |
| Other models | Tranception | 2022-05-27 | 7.2e+21 | 64 | |
| Other models | GPT-2 Medium (FlashAttention) | Stanford University, University at Buffalo | 2022-05-27 | 8.9e+20 | 8 |
| Other models | CogVideo | 2022-05-29 | NA | NA | |
| Other models | Diffusion-GAN | UT Austin, Microsoft | 2022-06-05 | NA | NA |
| Other models | DITTO | 2022-06-06 | 3.3e+18 | 8 | |
| Other models | MetaLM | Microsoft Research | 2022-06-13 | NA | NA |
| Other models | CoCa | Google Research | 2022-06-14 | 7.3e+22 | 2,048 |
| Other models | Parti | Google Research | 2022-06-22 | 5.1e+23 | NA |
| Other models | ProGen2-xlarge | 2022-06-27 | 1.4e+22 | NA | |
| Other models | Minerva (540B) | 2022-06-29 | 2.7e+24 | 1,024 | |
| Other models | CodeT5-large | Salesforce | 2022-07-05 | 2.7e+21 | NA |
| Other models | NLLB | Meta AI | 2022-07-06 | 1.8e+22 | NA |
| Other models | BLOOM-176B | Hugging Face, BigScience | 2022-07-11 | 3.7e+23 | 384 |
| Other models | ESM2-15B | 2022-07-21 | 7.4e+22 | 512 | |
| Other models | OmegaPLM | 2022-07-22 | 1.0e+22 | NA | |
| Other models | AlexaTM 20B | Amazon | 2022-08-02 | 2.0e+23 | 128 |
| Other models | GLM-130B | Tsinghua University | 2022-08-04 | 3.5e+23 | 768 |
| Other models | BlenderBot 3 | 2022-08-10 | 4.3e+23 | 128 | |
| Other models | BEIT-3 | Microsoft | 2022-08-22 | 7.0e+19 | NA |
| Other models | SauTech | 2022-09-14 | NA | 8 | |
| Other models | PaLI | 2022-09-14 | 1.7e+23 | 1,024 | |
| Other models | Whisper | OpenAI | 2022-09-21 | 4.2e+21 | NA |
| Other models | Make-A-Video | Meta AI | 2022-09-29 | NA | NA |
| Other models | DiffDock | Massachusetts Institute of Technology (MIT) | 2022-10-04 | 7.2e+19 | NA |
| Other models | AlphaTensor | DeepMind | 2022-10-05 | 7.1e+20 | 64 |
| Other models | Phenaki | 2022-10-05 | NA | NA | |
| Other models | GenSLM | 2022-10-11 | 1.4e+21 | NA | |
| Other models | Diplodocus | 2022-10-11 | NA | NA | |
| Other models | Flan-PaLM 540B | 2022-10-20 | 2.5e+24 | 512 | |
| Other models | LMSI-Palm | 2022-10-20 | NA | NA | |
| Other models | U-PaLM (540B) | 2022-10-20 | 2.5e+24 | 512 | |
| Other models | EnCodec | Meta AI | 2022-10-24 | NA | NA |
| Other models | eDiff-I | NVIDIA | 2022-11-02 | 5.5e+19 | NA |
| Other models | Mogrifier RLSTM (WT2) | DeepMind | 2022-11-03 | 1.4e+17 | NA |
| Other models | mT0-13B | Hugging Face, BigScience | 2022-11-03 | NA | NA |
| Other models | InternImage | 2022-11-10 | 2.4e+21 | NA | |
| Other models | AltCLIP_M9 | 2022-11-12 | NA | NA | |
| Other models | EVA-01 | 2022-11-14 | 1.5e+22 | 128 | |
| Other models | Galactica | Meta AI | 2022-11-16 | 3.2e+23 | 128 |
| Other models | Fusion in Encoder | Samsung | 2022-11-18 | 1.3e+20 | NA |
| Other models | AR-LDM | 2022-11-20 | 5.1e+20 | NA | |
| Other models | CICERO | Meta AI | 2022-11-22 | NA | NA |
| Other models | GPT-3.5 | OpenAI | 2022-11-28 | 2.6e+24 | NA |
| Other models | Discriminator Guidance | 2022-11-28 | 2.2e+20 | NA | |
| Other models | DiT-XL/2 + Discriminator Guidance | 2022-11-28 | NA | NA | |
| Other models | ALM 1.0 | 2022-11-28 | NA | NA | |
| Other models | GPT-3.5 Turbo | OpenAI | 2022-11-30 | NA | NA |
| Other models | DeepNash | DeepMind | 2022-12-01 | NA | NA |
| Other models | Vega v2 | 2022-12-04 | 7.8e+22 | 320 | |
| Other models | TranceptEve | University of Oxford, Harvard Medical School | 2022-12-10 | NA | NA |
| Other models | RT-1 | 2022-12-13 | NA | NA | |
| Other models | CaLM | University of Oxford | 2022-12-19 | 2.9e+19 | 4 |
| Other models | Hybrid H3-2.7B | Stanford University, University at Buffalo | 2022-12-28 | 6.5e+21 | 8 |
| Other models | VALL-E | Microsoft | 2023-01-05 | 1.0e+19 | NA |
| Other models | DreamerV3 | DeepMind, University of Toronto | 2023-01-10 | 2.2e+20 | 16 |
| Other models | Nucleotide Transformer | 2023-01-15 | 8.1e+21 | 128 | |
| Other models | Ankh_large | 2023-01-16 | 6.5e+21 | 64 | |
| Other models | MusicLM | 2023-01-26 | NA | NA | |
| Other models | DDPM-IP (CelebA) | Utrecht University | 2023-01-27 | 3.5e+20 | NA |
| Other models | BLIP-2 (Q-Former) | Salesforce Research | 2023-01-30 | 1.2e+21 | 16 |
| Other models | Gen-1 | Runway | 2023-02-06 | NA | NA |
| Other models | ProteinDT | 2023-02-09 | NA | NA | |
| Other models | ViT-22B | 2023-02-10 | 1.9e+23 | 1,024 | |
| Other models | BASIC-L + Lion | 2023-02-13 | NA | NA | |
| Other models | LLaMA-65B | Meta AI | 2023-02-24 | 5.5e+23 | 2,048 |
| Other models | DiT-XL/2 | 2023-03-02 | 6.0e+20 | NA | |
| Other models | AudioGen | Meta AI, Hebrew University of Jerusalem | 2023-03-05 | 9.5e+21 | NA |
| Other models | PaLM-E | Google, TU Berlin | 2023-03-06 | NA | NA |
| Other models | Claude | Anthropic | 2023-03-14 | NA | NA |
| Other models | LEP-AD | 2023-03-15 | NA | NA | |
| Other models | Falcon-40B | Technology Innovation Institute | 2023-03-15 | 2.4e+23 | 384 |
| Other models | PanGu-Σ | Huawei Noah's Ark Lab | 2023-03-20 | 4.7e+23 | 512 |
| Other models | Gen-2 | Runway | 2023-03-20 | NA | NA |
| Other models | Firefly | Adobe | 2023-03-21 | NA | NA |
| Other models | SigLIP 400M | Google DeepMind | 2023-03-27 | 4.9e+21 | 32 |
| Other models | VideoMAE V2 | 2023-03-29 | 9.7e+21 | 64 | |
| Other models | BloombergGPT | Bloomberg, Johns Hopkins University | 2023-03-30 | 2.4e+23 | 512 |
| Other models | Segment Anything Model | Meta AI | 2023-04-05 | 7.8e+21 | 256 |
| Other models | Incoder-6.7B | 2023-04-09 | 3.0e+21 | NA | |
| Other models | DINOv2 | Facebook AI Research, INRIA | 2023-04-14 | 7.4e+21 | NA |
| Other models | LLaVA | 2023-04-17 | 7.8e+22 | 8 | |
| Other models | Agile Soccer Robot | Google DeepMind | 2023-04-26 | NA | NA |
| Other models | StarCoder | 2023-05-09 | 8.5e+22 | 512 | |
| Other models | ImageBind | Meta AI | 2023-05-09 | NA | NA |
| Other models | PaLM 2 | 2023-05-10 | 7.3e+24 | NA | |
| Other models | InstructBLIP | 2023-05-11 | 1.9e+20 | 16 | |
| Other models | Med-PaLM 2 | Google Research, DeepMind | 2023-05-16 | NA | NA |
| Other models | CoEdiT-xxl | University of Minnesota, Grammarly | 2023-05-17 | NA | NA |
| Other models | ONE-PEACE | 2023-05-18 | 1.8e+20 | NA | |
| Other models | CodeT5+ | Salesforce | 2023-05-20 | NA | NA |
| Other models | Goat-7B | National University of Singapore | 2023-05-23 | NA | NA |
| Other models | PaLI-X | Google Research | 2023-05-29 | 5.6e+23 | NA |
| Other models | LTM-1 | Magic | 2023-06-06 | NA | NA |
| Other models | MusicGen | Meta AI | 2023-06-08 | NA | NA |
| Other models | HyenaDNA | 2023-06-27 | 1.8e+21 | 8 | |
| Other models | ERNIE 3.5 | Baidu | 2023-06-27 | NA | NA |
| Other models | Stable Diffusion XL (SDXL) | Stability AI | 2023-07-04 | NA | NA |
| Other models | Pangu-Weather | Huawei | 2023-07-05 | 4.0e+22 | 192 |
| Other models | xTrimoPGLM -100B | Tsinghua University, BioMap Research | 2023-07-06 | 6.2e+23 | 768 |
| Other models | InternLM | Shanghai AI Lab, SenseTime | 2023-07-06 | 1.0e+24 | NA |
| Other models | Claude 2 | Anthropic | 2023-07-11 | 3.9e+24 | NA |
| Other models | Llama 2-7B | Meta AI | 2023-07-18 | 8.4e+22 | NA |
| Other models | GPT3-2.7B (FlashAttention-2) | Stanford University, Princeton University | 2023-07-18 | NA | 8 |
| Other models | Llama 2-70B | Meta AI | 2023-07-18 | 8.1e+23 | 1,000 |
| Other models | AudioLM | Google Research | 2023-07-26 | 3.9e+18 | NA |
| Other models | RT-2 | Google DeepMind | 2023-07-28 | NA | NA |
| Other models | GGNN | 2023-08-05 | 7.6e+21 | 2 | |
| Other models | Qwen-VL | Alibaba | 2023-08-24 | NA | NA |
| Other models | PeptideBERT | Carnegie Mellon University (CMU) | 2023-08-28 | 4.9e+16 | 1 |
| Other models | Jais | 2023-08-29 | 4.9e+22 | 16 | |
| Other models | Swift | Intel Labs | 2023-08-30 | 5.3e+16 | 1 |
| Other models | Falcon-180B | Technology Innovation Institute | 2023-09-06 | 3.8e+24 | 4,096 |
| Other models | Robot Parkour | 2023-09-12 | NA | NA | |
| Other models | AlphaMissense | Google DeepMind | 2023-09-22 | NA | NA |
| Other models | GPT-4V | OpenAI | 2023-09-25 | NA | NA |
| Other models | Show-1 | National University of Singapore | 2023-09-27 | NA | NA |
| Other models | CTM (CIFAR-10) | Stanford University, Sony | 2023-10-01 | NA | 4 |
| Other models | FinGPT-13B | 2023-10-07 | 1.6e+23 | 1 | |
| Other models | RoseTTAFold All-Atom (RFAA) | 2023-10-09 | 2.1e+20 | NA | |
| Other models | Ferret (13B) | Columbia University, Apple | 2023-10-11 | NA | 8 |
| Other models | RT-2-X | Google DeepMind | 2023-10-13 | NA | NA |
| Other models | ERNIE 4.0 | Baidu | 2023-10-17 | NA | NA |
| Other models | PaLI-3 | 2023-10-17 | NA | NA | |
| Other models | DALL·E 3 | OpenAI | 2023-10-19 | NA | NA |
| Other models | CODEFUSION (Python) | Microsoft, Microsoft Research | 2023-10-26 | 7.9e+18 | 4 |
| Other models | DiT-XL/2 + CADS | ETH Zurich, Disney Research | 2023-10-26 | NA | NA |
| Other models | ChatGLM3-6B | Zhipu AI | 2023-10-27 | 5.0e+22 | NA |
| Other models | Skywork-13B | Kunlun Inc. | 2023-10-30 | 2.5e+23 | 512 |
| Other models | Cohere Embed | Cohere | 2023-11-02 | NA | NA |
| Other models | Yi-34B | 01.AI | 2023-11-02 | 6.1e+23 | 128 |
| Other models | RT-Trajectory | 2023-11-03 | NA | NA | |
| Other models | BLUUMI | University of Turku, Hugging Face | 2023-11-03 | NA | NA |
| Other models | Grok-1 | xAI | 2023-11-04 | 2.9e+24 | NA |
| Other models | LLaVA 1.5 | 2023-11-05 | 7.8e+22 | 8 | |
| Other models | CogVLM-17B | 2023-11-06 | 6.3e+22 | NA | |
| Other models | GPT-4 Turbo | OpenAI | 2023-11-06 | 2.2e+25 | NA |
| Other models | OmniVec | TensorTour | 2023-11-07 | NA | NA |
| Other models | mPLUG-Owl2 | Alibaba | 2023-11-07 | NA | NA |
| Other models | MultiBand Diffusion | 2023-11-08 | 2.6e+19 | NA | |
| Other models | SPHINX (Llama 2 13B) | 2023-11-13 | 3.0e+22 | 32 | |
| Other models | Volcano 13B | 2023-11-13 | 4.6e+22 | NA | |
| Other models | GraphCast | Google DeepMind | 2023-11-14 | 2.1e+22 | NA |
| Other models | Qwen-Audio-Chat | Alibaba | 2023-11-14 | NA | NA |
| Other models | Nemotron-3-8B | NVIDIA | 2023-11-15 | 1.8e+23 | 1,024 |
| Other models | AndesGPT | Oppo Mobile Telecommunications | 2023-11-16 | NA | NA |
| Other models | Claude 2.1 | Anthropic | 2023-11-21 | NA | NA |
| Other models | Inflection-2 | Inflection AI | 2023-11-22 | 1.0e+25 | 5,000 |
| Other models | GNoME for crystal discovery | Google DeepMind | 2023-11-29 | NA | 4 |
| Other models | PPLX-70B-Online | Perplexity | 2023-11-29 | NA | NA |
| Other models | Qwen-72B | Alibaba | 2023-11-30 | 1.3e+24 | NA |
| Other models | Mamba-24M (SC09) | 2023-12-01 | NA | NA | |
| Other models | Gemini 1.0 Pro | Google DeepMind | 2023-12-06 | 1.8e+24 | NA |
| Other models | Llama Guard | Meta AI | 2023-12-07 | 1.6e+23 | NA |
| Other models | SeamlessM4T | 2023-12-08 | NA | NA | |
| Other models | W.A.L.T | 2023-12-11 | NA | NA | |
| Other models | Mixtral 8x7B | Mistral AI | 2023-12-11 | 7.7e+23 | NA |
| Other models | VILA-13B | 2023-12-12 | 2.3e+21 | 128 | |
| Other models | FunSearch | Google DeepMind | 2023-12-14 | 3.9e+23 | NA |
| Other models | CogAgent | Tsinghua University, Zhipu AI | 2023-12-14 | 6.7e+22 | NA |
| Other models | Gemini Nano-2 | Google DeepMind | 2023-12-19 | NA | NA |
| Other models | Gemini Nano-1 | Google DeepMind | 2023-12-19 | NA | NA |
| Other models | nekomata-14b | rinna | 2023-12-21 | 2.6e+23 | 256 |
| Other models | CoRe | Tsinghua University | 2023-12-29 | NA | NA |
| Other models | Kimi Explorer | Moonshot | 2024-01-01 | NA | NA |
| Other models | Palmyra X 003 | Writer | 2024-01-01 | NA | NA |
| Other models | AlphaGeometry | Google DeepMind, New York University (NYU) | 2024-01-17 | NA | NA |
| Other models | Qwen-VL-Max | Alibaba | 2024-01-25 | NA | NA |
| Other models | Qwen1.5-72B | Alibaba | 2024-02-04 | 1.3e+24 | NA |
| Other models | Aya | 2024-02-12 | NA | 128 | |
| Other models | Gemini 1.5 Pro | Google DeepMind | 2024-02-15 | 1.6e+25 | NA |
| Other models | Sora Turbo | OpenAI | 2024-02-15 | NA | NA |
| Other models | Sora | OpenAI | 2024-02-15 | NA | NA |
| Other models | Stable Diffusion 3 | Stability AI | 2024-02-22 | 5.0e+22 | NA |
| Other models | MegaScale (Production) | ByteDance, Peking University | 2024-02-23 | 3.9e+24 | 12,288 |
| Other models | Mistral Large | Mistral AI | 2024-02-26 | 1.1e+25 | NA |
| Other models | Claude 3 Sonnet | Anthropic | 2024-03-04 | NA | NA |
| Other models | Claude 3 Opus | Anthropic | 2024-03-04 | 1.6e+25 | NA |
| Other models | Aramco Metabrain AI | Saudi Aramco | 2024-03-04 | 1.0e+25 | NA |
| Other models | Inflection-2.5 | Inflection AI | 2024-03-07 | 8.0e+24 | NA |
| Other models | ManiGaussian | 2024-03-13 | NA | 2 | |
| Other models | MM1-30B | Apple | 2024-03-14 | 4.9e+23 | NA |
| Other models | DBRX | Databricks | 2024-03-27 | 2.6e+24 | NA |
| Other models | ReALM | Apple | 2024-03-29 | NA | NA |
| Other models | Reka Core | Reka AI | 2024-04-15 | 8.4e+24 | NA |
| Other models | Llama 3-70B | Meta AI | 2024-04-18 | 7.9e+24 | NA |
| Other models | GenCast | Google DeepMind | 2024-05-01 | 8.2e+20 | 32 |
| Other models | VILA1.5-13B | 2024-05-03 | 2.3e+21 | 128 | |
| Other models | AlphaFold 3 | Google DeepMind, Isomorphic Labs | 2024-05-08 | 4.1e+22 | 256 |
| Other models | GPT-4o | OpenAI | 2024-05-13 | 3.8e+25 | NA |
| Other models | Yi-Large | 01.AI | 2024-05-13 | 1.8e+24 | NA |
| Other models | Octo-Base | 2024-05-20 | 5.8e+20 | 128 | |
| Other models | GLM-4 (0520) | Zhipu AI | 2024-05-20 | NA | NA |
| Other models | ALLaM 7B | 2024-05-21 | 9.0e+22 | NA | |
| Other models | ALLaM adapted 70B | 2024-05-21 | 1.1e+24 | NA | |
| Other models | Qwen2-72B | Alibaba | 2024-06-07 | 3.0e+24 | NA |
| Other models | Llama-3.1-Nemotron-70B-Instruct | NVIDIA, Meta AI | 2024-06-12 | 7.9e+24 | NA |
| Other models | OpenVLA | 2024-06-13 | 1.1e+23 | 64 | |
| Other models | Nemotron-4 340B | NVIDIA | 2024-06-14 | 1.8e+25 | 6,144 |
| Other models | DeepSeek-Coder-V2 236B | DeepSeek | 2024-06-17 | 1.3e+24 | NA |
| Other models | Claude 3.5 Sonnet | Anthropic | 2024-06-20 | 2.7e+25 | NA |
| Other models | Cambrian-1-34B | New York University (NYU) | 2024-06-24 | NA | 512 |
| Other models | ESM3 (98B) | 2024-06-25 | 1.1e+24 | NA | |
| Other models | Ernie 4.0 Turbo | Baidu | 2024-06-28 | NA | NA |
| Other models | SenseChat 5.5 | SenseTime | 2024-07-06 | NA | NA |
| Other models | Mathstral | Mistral AI | 2024-07-16 | NA | NA |
| Other models | DeepL LLM | DeepL | 2024-07-16 | NA | NA |
| Other models | GPT-4o mini | OpenAI | 2024-07-18 | 7.4e+24 | NA |
| Other models | Mistral Large 2 | Mistral AI | 2024-07-24 | 2.1e+25 | NA |
| Other models | AFM-server | Apple | 2024-07-29 | 4.3e+24 | 8,192 |
| Other models | AFM-on-device | Apple | 2024-07-29 | 4.5e+23 | 2,048 |
| Other models | LLaVA-OV-72B | 2024-08-06 | 3.0e+24 | NA | |
| Other models | Table Tennis Agent | Google DeepMind | 2024-08-07 | NA | NA |
| Other models | Grok-2 | xAI | 2024-08-13 | 3.0e+25 | NA |
| Other models | Jamba 1.5-Large | AI21 Labs | 2024-08-22 | NA | NA |
| Other models | Hairuo | Inspur | 2024-08-29 | NA | NA |
| Other models | GLM-4-Plus | Zhipu AI | 2024-08-29 | 3.6e+25 | NA |
| Other models | AlphaProteo | Google DeepMind | 2024-09-05 | NA | NA |
| Other models | Hunyuan Turbo | Tencent | 2024-09-05 | NA | NA |
| Other models | DeepSeek-V2.5 | DeepSeek | 2024-09-06 | 1.8e+24 | NA |
| Other models | o1-mini | OpenAI | 2024-09-12 | NA | NA |
| Other models | o1-preview | OpenAI | 2024-09-12 | NA | NA |
| Other models | Qwen2.5-32B | Alibaba | 2024-09-17 | 3.5e+24 | NA |
| Other models | Qwen2.5 Instruct (72B) | Alibaba | 2024-09-19 | 7.9e+24 | NA |
| Other models | Qwen2.5-72B | Alibaba | 2024-09-19 | 7.8e+24 | NA |
| Other models | Oryx 34B | 2024-09-19 | NA | 64 | |
| Other models | Telechat2-115B | China Telecom | 2024-09-20 | 6.9e+24 | NA |
| Other models | PixelDance | ByteDance | 2024-09-24 | NA | NA |
| Other models | Llama 3.2 11B | Meta AI | 2024-09-24 | 5.8e+23 | NA |
| Other models | Movie Gen Video | Meta AI | 2024-10-04 | 1.6e+24 | 6,144 |
| Other models | GR-2 | ByteDance | 2024-10-08 | NA | NA |
| Other models | Palmyra X 004 | Writer | 2024-10-09 | NA | NA |
| Other models | RDT-1B | Tsinghua University | 2024-10-10 | 4.1e+22 | 48 |
| Other models | CHAI-1 | Chai discovery | 2024-10-15 | 7.8e+21 | 128 |
| Other models | Yi-Lightning | 01.AI | 2024-10-18 | 1.5e+24 | 2,000 |
| Other models | NVLM-D 72B | NVIDIA | 2024-10-22 | 3.0e+24 | 128 |
| Other models | NVLM-H 72B | NVIDIA | 2024-10-22 | 3.0e+24 | 128 |
| Other models | NVLM-X 72B | NVIDIA | 2024-10-22 | 3.0e+24 | 128 |
| Other models | Doubao-pro | ByteDance | 2024-10-28 | 2.5e+25 | NA |
| Other models | Hunyuan-Large | Tencent | 2024-11-06 | 3.5e+24 | NA |
| Other models | SeedEdit | ByteDance | 2024-11-11 | NA | NA |
| Other models | Gemini-Exp-1114 | Google DeepMind | 2024-11-15 | NA | NA |
| Other models | k0-math | Moonshot | 2024-11-16 | NA | NA |
| Other models | Pixtral Large | Mistral AI | 2024-11-18 | NA | NA |
| Other models | Suno v4 | Suno | 2024-11-19 | NA | NA |
| Other models | Fugatto 1 | NVIDIA | 2024-11-25 | NA | 32 |
| Other models | Amazon Nova Pro | Amazon | 2024-12-03 | 6.0e+24 | NA |
| Other models | Infinity | ByteDance | 2024-12-05 | NA | NA |
| Other models | NVILA 15B | 2024-12-05 | NA | 128 | |
| Other models | o1 | OpenAI | 2024-12-05 | NA | NA |
| Other models | Llama 3.3 70B | Meta AI | 2024-12-06 | 6.9e+24 | NA |
| Other models | EXAONE 3.5 32B | LG AI Research | 2024-12-09 | 1.3e+24 | NA |
| Other models | Gemini 2.0 Pro | Google DeepMind | 2024-12-11 | NA | NA |
| Other models | Apollo 7B | Meta AI, Stanford University | 2024-12-13 | NA | 128 |
| Other models | Veo 2 | Google DeepMind | 2024-12-16 | NA | NA |
| Other models | o3 | OpenAI | 2024-12-20 | NA | NA |
| Other models | DeepSeek-V3 | DeepSeek | 2024-12-24 | 3.4e+24 | 2,048 |
| Other models | STORM-B/8 | 2024-12-31 | NA | 16 | |
| Other models | Stable Point Aware 3D (SPAR3D) | 2025-01-08 | NA | NA | |
| Other models | INTELLECT-MATH | Prime Intellect | 2025-01-17 | NA | NA |
| Other models | Eagle 2 | 2025-01-20 | 4.7e+22 | 256 | |
| Other models | DeepSeek-R1 | DeepSeek | 2025-01-20 | 4.0e+24 | NA |
| Other models | Kimi k1.5 | Moonshot | 2025-01-22 | NA | NA |
| Other models | Doubao-1.5-pro | ByteDance | 2025-01-22 | NA | NA |
| Other models | Computer-Using Agent (CUA) | OpenAI | 2025-01-23 | NA | NA |
| Other models | o3-mini | OpenAI | 2025-01-31 | NA | NA |
| Other models | Eurus-2-7B-PRIME | 2025-02-03 | NA | NA | |
| Other models | Claude 3.7 Sonnet | Anthropic | 2025-02-24 | 3.3e+25 | NA |
| Other models | Mercury | Inception Labs | 2025-02-27 | NA | NA |
| Other models | GPT-4.5 | OpenAI | 2025-02-27 | 2.1e+26 | NA |
| Other models | QwQ-32B | Alibaba | 2025-03-06 | 3.5e+24 | NA |
| Other models | Mistral OCR | Mistral AI | 2025-03-06 | NA | NA |
| Other models | Hunyuan-TurboS | Tencent | 2025-03-11 | NA | NA |
| Other models | EXAONE Deep 32B | LG AI Research | 2025-03-16 | 1.3e+24 | 512 |
| Other models | ERNIE-4.5-VL-424B-A47B (文心大模型4.5) | Baidu | 2025-03-16 | NA | NA |
| Other models | Diffusion Renderer | 2025-03-22 | NA | 32 | |
| Other models | Gemini 2.5 Pro | Google DeepMind | 2025-03-25 | NA | NA |
| Other models | Llama 4 Maverick | Meta AI | 2025-04-05 | 2.2e+24 | NA |
| Other models | Llama 4 Scout | Meta AI | 2025-04-05 | 4.1e+24 | NA |
| Other models | Pangu Ultra | Huawei | 2025-04-10 | 1.1e+25 | 8,192 |
| Other models | gpt-image-1 | OpenAI | 2025-04-23 | NA | NA |
| Other models | Qwen3-235B-A22B | Alibaba | 2025-04-29 | 4.8e+24 | NA |
| Other models | Seed1.5-VL | ByteDance | 2025-05-11 | 1.4e+24 | NA |
| Other models | Imagen 4 | 2025-05-20 | NA | NA | |
| Other models | Veo 3 | Google DeepMind | 2025-05-21 | NA | NA |
| Other models | Claude Sonnet 4 | Anthropic | 2025-05-22 | NA | NA |
| Other models | Claude Opus 4 | Anthropic | 2025-05-22 | NA | NA |
| Other models | Qwen3 Embedding | Alibaba | 2025-06-05 | NA | NA |
| Other models | Seed-1.6-Thinking | ByteDance | 2025-06-11 | NA | NA |
| Other models | FGN | Google DeepMind | 2025-06-12 | 9.6e+21 | NA |
| Other models | Kimi K2 | Moonshot | 2025-07-11 | 3.0e+24 | NA |
| Other models | Gemini Embedding | Google DeepMind | 2025-07-14 | NA | NA |
| Other models | EXAONE 4.0 (32B) | LG AI Research | 2025-07-15 | 2.7e+24 | 512 |
| Other models | Qwen3-Coder-480B-A35B | Alibaba | 2025-07-22 | 1.6e+24 | NA |
| Other models | Gemini 2.5 Deep Think | Google, Google DeepMind | 2025-08-01 | NA | NA |
| Other models | MindLink-72B | Kunlun Inc. | 2025-08-01 | NA | NA |
| Other models | Qwen Image | Alibaba | 2025-08-04 | NA | NA |
| Other models | GLM 4.5 | Zhipu AI, Tsinghua University | 2025-08-05 | 4.4e+24 | NA |
| Other models | gpt-oss-120b | OpenAI | 2025-08-05 | 4.9e+24 | NA |
| Other models | gpt-oss-20b | OpenAI | 2025-08-05 | 5.5e+23 | NA |
| Other models | GPT-5 | OpenAI | 2025-08-07 | NA | NA |
| Other models | GPT-5 mini | OpenAI | 2025-08-07 | NA | NA |
| Other models | GPT-5 nano | OpenAI | 2025-08-07 | NA | NA |
| Other models | Qwen3-Max | Alibaba | 2025-09-05 | 1.5e+25 | NA |
Epoch's work is free to use, distribute, and reproduce provided the source and authors are credited under the Creative Commons BY license.
Download this data
CSV, Updated Nov. 24, 2025
Explore this data
Key data on over 170 AI accelerators, such as graphics processing units (GPUs) and tensor processing units (TPUs).