Closed-weight language models in 2021
14
Models
12 organisations
0%
Open weights
0 open, 14 closed
7
Frontier
By Epoch's flag
1.2T
Largest traceable
GLaM
2021
Span
10 with a traceable size
Releases
by monthOrganisation
Records
14| Model | Organisation | Released | Parameters | Weights |
|---|---|---|---|---|
| ERNIE 3.0 Titan | Baidu | 2021-12 | 260B | Closed |
| GLaM | 2021-12 | 1.2T | Closed | |
| Gopher (280B) | DeepMind | 2021-12 | 280B | Closed |
| T-NLRv5 XXL | Microsoft | 2021-12 | 5B | Closed |
| Megatron-Turing NLG 530B | Microsoft | 2021-10 | 530B | Closed |
| Yuan 1.0 | Inspur | 2021-10 | 246B | Closed |
| FLAN 137B | Google Research | 2021-09 | 137B | Closed |
| HyperCLOVA 204B | NAVER | 2021-09 | 204B * | Closed |
| PermuteFormer | Peking University | 2021-09 | 150M * | Closed |
| Turing ULRv5 | Microsoft | 2021-09 | 2B | Closed |
| Jurassic-1-Jumbo | AI21 Labs | 2021-08 | 178B | Closed |
| Codex | OpenAI | 2021-07 | 12B * | Closed |
| Adaptive Input Transformer + RD | Microsoft Research Asia | 2021-06 | 247M * | Closed |
| PLUG | Alibaba | 2021-04 | 27B | Closed |
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.