// RELEASE TIMELINE

Models Released in 2017

Browse every generative AI model released in 2017 — large language models, image generators, code models and more — ordered by release date. Compare their specs and rankings on our benchmarks page, or put any two side by side with compare. See also models from 2016 and models from 2018.

79 models

PixelSNAIL (CIFAR 10)

🇺🇸 University of California (UC) Berkeley

Dec 2017
Language Open (Restricted)

PixelSNAIL (ImageNet)

🇺🇸 University of California (UC) Berkeley

Dec 2017
Language Open (Restricted)

Tacotron 2

🇺🇸 Google

Dec 2017
Speech Proprietary

WGAN (Wasserstein GAN)

🇺🇸 Facebook AI Research

Dec 2017
Language Proprietary

2-layer-LSTM+Deep-Gradient-Compression

🇨🇳 Tsinghua University

Dec 2017 6M
Language Proprietary

AlphaZero

🇬🇧 DeepMind

Dec 2017
Multimodal Proprietary

PNASNet-5

🇺🇸 Johns Hopkins University

Dec 2017 86.1M
Vision Proprietary

DL scaling Image

🇨🇳 Baidu

Dec 2017 121M
Vision Proprietary

DL scaling LM

🇨🇳 Baidu

Dec 2017 177M
Language Proprietary

DL scaling speech

🇨🇳 Baidu

Dec 2017 193M
Speech Proprietary

TriNet

🇩🇪 Visual Computing Institute

Nov 2017
Video Proprietary

AWD-LSTM-MoS + dynamic evaluation (PTB, 2017)

🇺🇸 Carnegie Mellon University (CMU)

Nov 2017 22M
Language Proprietary

AWD-LSTM-MoS + dynamic evaluation (WT2, 2017)

🇺🇸 Carnegie Mellon University (CMU)

Nov 2017 35M
Language Proprietary

VQ-VAE

🇬🇧 DeepMind

Nov 2017
Vision Proprietary

DCN+

🇺🇸 Salesforce Research

Oct 2017
Language Proprietary

Fraternal dropout + AWD-LSTM 3-layer (PTB)

🇺🇸 University of Montreal / Université de Montréal

Oct 2017 24M
Language Proprietary

Fraternal dropout + AWD-LSTM 3-layer (WT2)

🇺🇸 Jagiellonian University

Oct 2017 34M
Language Proprietary

S-Norm

🇺🇸 University of Washington

Oct 2017
Language Proprietary

PhraseCond

🇺🇸 Carnegie Mellon University (CMU)

Oct 2017
Language Proprietary

ProgressiveGAN

🇺🇸 NVIDIA

Oct 2017
Vision Proprietary

LRSO-GAN

🇨🇳 University of Technology Sydney

Oct 2017
Vision Proprietary

AlphaGo Master

🇬🇧 DeepMind

Oct 2017
Multimodal Proprietary

AlphaGo Zero

🇬🇧 DeepMind

Oct 2017 46.4M
Multimodal Proprietary

Rainbow DQN

🇬🇧 DeepMind

Oct 2017
Multimodal Proprietary

AWD-LSTM+WT+Cache+IOG (PTB)

🇯🇵 NTT Communication Science Laboratories

Sep 2017 30M
Language Proprietary

AWD-LSTM+WT+Cache+IOG (WT2)

🇯🇵 NTT Communication Science Laboratories

Sep 2017 53M
Language Proprietary

AWD-LSTM + dynamic eval (PTB)

🇺🇸 University of Edinburgh

Sep 2017 24M
Language Proprietary

AWD-LSTM + dynamic eval (WT2)

🇺🇸 University of Edinburgh

Sep 2017 33M
Language Proprietary

LSTM + dynamic eval

🇺🇸 University of Edinburgh

Sep 2017 50M
Language Proprietary

ISS

🇺🇸 Duke University

Sep 2017 11.1M
Language Proprietary

PyramidNet

🇰🇷 Korea Advanced Institute of Science and Technology (KAIST)

Sep 2017 26M
Vision Open (Restricted)

GL-LWGC-AWD-MoS-LSTM + dynamic evaluation (PTB)

🇮🇱 Ben-Gurion University

Aug 2017 26M
Language Proprietary

GL-LWGC-AWD-MoS-LSTM + dynamic evaluation (WT2)

🇺🇸 Ben-Gurion University of the Negev

Aug 2017 38M
Language Proprietary

D-LSRC(100)+KN5 (PTB)

🇺🇸 Saarland University

Aug 2017 6M
Language Proprietary

GRU + p-tHSM (pretrain via Brown) (PTB)

🇨🇳 Beihang University

Aug 2017
Language Proprietary

GRU + p-tHSM (pretrain via Brown) (WT2)

🇨🇳 Beihang University

Aug 2017
Language Proprietary

Libratus

🇺🇸 Carnegie Mellon University (CMU)

Aug 2017
Multimodal Proprietary

Adversarial Joint Adaptation Network (ResNet)

🇨🇳 Tsinghua University

Aug 2017 60M
Vision Proprietary

NeuMF (Pinterest)

🇨🇳 Shandong University

Aug 2017
Language Proprietary

Cutout-regularized net

🇨🇦 University of Guelph

Aug 2017
Vision Proprietary

EI-REHN-1000D

🇰🇷 Korea Advanced Institute of Science and Technology (KAIST)

Aug 2017 19M
Language Proprietary

EI-REHN-1200D (PTB)

🇰🇷 Korea Advanced Institute of Science and Technology (KAIST)

Aug 2017 25M
Language Proprietary

OpenAI TI7 DOTA 1v1

🇺🇸 OpenAI

Aug 2017
Multimodal Proprietary

AWD-LSTM - 3-layer LSTM (tied) + continuous cache pointer (PTB)

🇺🇸 Salesforce Research

Aug 2017 24M
Language Proprietary

AWD-LSTM - 3-layer LSTM (tied) + continuous cache pointer (WT2)

🇺🇸 Salesforce Research

Aug 2017 33M
Language Proprietary

RetinaNet-R101

🇺🇸 Facebook AI Research

Aug 2017 53M
Vision Proprietary

GSM

🇨🇳 Peking University

Jul 2017
Language Proprietary

ConvS2S (ensemble of 8 models)

🇺🇸 Meta AI

Jul 2017
Language Proprietary

PSPNet

🇺🇸 Chinese University of Hong Kong (CUHK)

Jul 2017
Vision Proprietary

4 layer Densely Connected LSTM 14M (PTB)

🇧🇪 Ghent University

Jul 2017 14M
Language Proprietary

Densely Connected LSTM + Var. Dropout

🇧🇪 Ghent University

Jul 2017 23M
Language Proprietary

AWD-LSTM

🇬🇧 DeepMind

Jul 2017 24M
Language Proprietary

JFT

🇺🇸 Google Research

Jul 2017 44.7M
Vision Proprietary

DeepLoc

🇩🇰 Technical University of Denmark

Jul 2017
Language Proprietary

NoisyNet-Dueling

🇬🇧 DeepMind

Jun 2017
Multimodal Proprietary

DeepLabV3

🇺🇸 Google

Jun 2017
Vision Proprietary

HRA

🇺🇸 Maluuba

Jun 2017
Multimodal Proprietary

Transformer

🇺🇸 Google Research

Jun 2017 213M
Language Proprietary

EDSR

🇺🇸 Seoul National University

Jun 2017
Vision Proprietary

Reading Twice for NLU

🇬🇧 DeepMind

Jun 2017
Language Proprietary

2017 was another fast-moving year for generative AI. The models listed above span large language models (LLMs), image and video generators, code models and more, each with its own parameter count, context window, licensing and availability. To understand how the models of 2017 actually perform, compare their scores on our benchmarks page, and use compare to evaluate any two releases side by side.

Tracking AI by release year makes it easy to see how the frontier moves. Browse AI models from 2016 or AI models from 2018 to compare how capabilities, context windows and open-weights availability evolved year over year.