Meta
29 July 2004
FAIR under LeCun from 2013. PyTorch, and the Llama models that seeded the open-weights ecosystem.
American technology company
Also: Meta AI, Facebook AI Research, FAIR, Facebook
Connections
Developed 86
- DeepFace2014
- YOLO2015
- Character-enriched word2vec2016
- DenseNet-2642016
- DTN (Domain Transfer Network)2016
- ResNeXt-101 (64×4d)2016
- ResNeXt-502016
- Mask R-CNN2017
- ConvS2S (ensemble of 8 models)2017
- RetinaNet-R1012017
- RetinaNet-R502017
- ResNeXt-101 32x48d2018
- Big Transformer for Back-Translation2018
- Transformer (Adaptive Input Embeddings) WT1032018
- XLM2019
- LaNet-L (CIFAR-10)2019
- RoBERTa Large2019
- Pluribus2019
- Adaptive Inputs + LayerDrop2019
- AlphaX-12019
- Base LM + kNN LM + Continuous Cache2019
- CamemBERT2019
- Sandwich Transformer2019
- Transformer - LibriVox + Decoding/Rescoring2019
- DD-PPO2019
- Feedback Transformer2020
- Retrieval-Augmented Generator2020
- DETR2020
- SemExp2020
- DeLighT2020
- wave2vec 2.0 LARGE2020
- ESM1b2020
- DeiT-B2021
- MSA Transformer2021
- Generative BST2021
- HuBERT2021
- SEER2021
- 6-Act Tether2021
- XLMR-XXL2021
- Masked Autoencoders ViT-H2021
- Contriever2021
- XGLM-7.5B2021
- Detic2022
- data2vec (language)2022
- data2vec (speech)2022
- data2vec (vision)2022
- ViT-G (model soup)2022
- Make-A-Scene2022
- Sparse all-MLP2022
- OPT-175B2022
- NLLB2022
- ESM2-15B2022
- BlenderBot 32022
- Make-A-Video2022
- Diplodocus2022
- EnCodec2022
- Galactica2022
- CICERO2022
- LLaMA2023
- AudioGen2023
- Segment Anything Model2023
- Incoder-6.7B2023
- DINOv22023
- ImageBind2023
- MusicGen2023
- Llama 2 70B2023
- Llama 2-7B2023
- MultiBand Diffusion2023
- Llama Guard2023
- SeamlessM4T2023
- Llama 3-70B2024
- Llama 3.1 405B2024
- Llama 3.2 11B2024
- Movie Gen Video2024
- Llama-3.1-Nemotron-70B-Instruct2024
- Llama 3.3 70B2024
- Apollo 7B2024
- Llama 4 Behemoth (preview)2025
- Llama 4 Maverick2025
- Llama 4 Scout2025
- Meta's Generative Ads Model (GEM)2025
- Muse Spark2026
- Muse Spark 1.22026
- Muse Spark 1.32026
- Muse Wonder2026
- Muse Wonder Full2026
Sources: AI4U editorial, Wikidata.