# vit-base-patch16-224-in21k by Google: Open-Weight Model Source: https://savrn.com/models/vit-base-patch16-224-in21k Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## Runs On What it takes to serve vit-base-patch16-224-in21k (86M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them. | Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits | | --- | --- | --- | --- | --- | --- | | 16-bit | 0.2 GB | 0.2 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 8-bit | 0.1 GB | 0.1 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 4-bit | 0.0 GB | 0.1 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the [SAVRN Index](https://savrn.com/ai-index/pricing/gpus), read Oct 24, 2026. [vit-base-patch16-224-in21k on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/vit-base-patch16-224-in21k/gpus) ## SAVRN's Notes on vit-base-patch16-224-in21k Released in March 2022, this one does image feature extraction: 86M parameters, pre-trained on ImageNet-21k at 224 by 224, 14 million images across 21,843 classes. Memory is a non-issue at 0.2 GB in 16-bit, and 4-bit weights round to 0.0 GB, so skip quantizing. The cheapest Index setup, one MI300X with 192 GB at $1.85 an hour on demand, bills by the hour, so throughput, not fit, is the whole question. Apache 2.0 covers commercial use, modification and redistribution with a patent grant; keep the notices and state significant changes. There is no context length and no host token price on the Index, since it produces features, not tokens. The weights ship in safetensors, pytorch, jax and tf, and the releasing team did not write the model card, so the paper, An Image is Worth 16x16 Words, is the reference to read before you wire it in. SAVRN Research, 2026-09-18 ## Model Card By Google, published under apache-2.0, revision b4569560a39a. ### Vision Transformer (base-sized model) Vision Transformer (ViT) model pre-trained on ImageNet-21k (14 million images, 21,843 classes) at resolution 224x224. It was introduced in the paper [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://savrn.com/papers/an-image-is-worth-16x16-words-transformers-for-image-recognition-at-scale) by Dosovitskiy et al. and first released in [this repository](https://github.com/google-research/vision_transformer). However, the weights were converted from the [timm repository](https://github.com/rwightman/pytorch-image-models) by Ross Wightman, who already converted the weights from JAX to PyTorch. Credits go to him. Disclaimer: The team releasing ViT did not write a model card for this model so this model card has been written by the Hugging Face team. ### Model description The Vision Transformer (ViT) is a transformer encoder model (BERT-like) pretrained on a large collection of images in a supervised fashion, namely ImageNet-21k, at a resolution of 224x224 pixels. [Read the full model card (659 words)](https://savrn.com/models/vit-base-patch16-224-in21k/card) ## Configuration Architecture ViTModel Layers 12 Hidden size 768 Feed-forward size 3,072 Attention heads 12 Model type vit ## Identity and Version Repository google/vit-base-patch16-224-in21k Publisher Google Task Image feature extraction Modality Other Library transformers Parameters 86M parameters Languages tf, jax, vit Revision b4569560a39a0f1af58e3ddaf17facf20ab919b0 First published 2022-03-02 Last updated 2024-02-05 ## Files and Weights 8 files, 1.4 GB in total. The weights are 4 files totalling 1.4 GB in bin, h5, msgpack, safetensors. Weights4 files · 1.4 GB Configuration2 files · 662 B Documentation1 file · 5.4 KB Repository1 file · 744 B Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | flax_model.msgpack | Weights | 345.6 MB | aa00da5f7abb | | model.safetensors | Weights | 345.6 MB | fd4e1169c7aa | | pytorch_model.bin | Weights | 345.6 MB | 84066da0f5d8 | | tf_model.h5 | Weights | 345.8 MB | 1f76b72dc4a8 | | config.json | Configuration | 502 B | — | | preprocessor_config.json | Configuration | 160 B | — | | README.md | Documentation | 5.4 KB | — | | .gitattributes | Repository | 744 B | — | ## License and Download License apache-2.0 Access Open weights, no gate Download size 1.4 GB [Download from Google](https://huggingface.co/google/vit-base-patch16-224-in21k) Released by Google through its official repository on Hugging Face. [Read the license](https://www.apache.org/licenses/LICENSE-2.0). ## Built From - Described by [arXiv:2006.03677](https://savrn.com/papers/visual-transformers-token-based-image-representation-and-processing-for-computer-vision) - Described by [arXiv:2010.11929](https://savrn.com/papers/an-image-is-worth-16x16-words-transformers-for-image-recognition-at-scale) - Trained on (disclosed) imagenet-21k ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 1.4 GB | | 16-bit | 0.2 GB | | 8-bit | 0.1 GB | | 4-bit | 0.0 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Built on This Model - Derived from[fairface_age_image_detection](https://savrn.com/models/fairface-age-image-detection) - Derived from[rorshark-vit-base](https://savrn.com/models/rorshark-vit-base) - Derived from[nsfw-classifier](https://savrn.com/models/nsfw-classifier) ## Compare vit-base-patch16-224-in21k - [dinov2-base vs vit-base-patch16-224-in21k](https://savrn.com/models/compare/dinov2-base-vs-vit-base-patch16-224-in21k) - [dinov2-small vs vit-base-patch16-224-in21k](https://savrn.com/models/compare/dinov2-small-vs-vit-base-patch16-224-in21k) ## Questions About vit-base-patch16-224-in21k ### How much GPU memory does vit-base-patch16-224-in21k need? About 0.2 GB at 16-bit and 0.1 GB at 4-bit: the weights (86M parameters) plus a working margin. A long context needs more. ### What is the cheapest GPU to run vit-base-patch16-224-in21k on? At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists. ### Can I use vit-base-patch16-224-in21k commercially? Yes. vit-base-patch16-224-in21k is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors. ## Similar Models Model · Image feature extraction ### [dinov2-base](https://savrn.com/models/dinov2-base) [AI at Meta](https://savrn.com/model-publishers/facebook) Vision Transformer (ViT) model trained using the DINOv2 method. It was introduced in the paper DINOv2: Learning Robust Visual Features without Supervision by Oquab et al. and first released in this repository. Disclaimer: The team releasing DINOv2 did not write a model card for this model so this model card has been written by the Hugging Face team. The Vision Transformer (ViT) is a transformer encoder model (BERT-like) pretrained on a large collection of images in a self-supervised fashion. Images are presented to the model as a sequence of fixed-size patches, which are linearly embedded. One also adds a [CLS] token to the beginning of a sequence to use it for classification tasks. One… Open weights apache-2.0 87M parameters transformers [View model](https://savrn.com/models/dinov2-base) Model · Image feature extraction ### [dinov2-small](https://savrn.com/models/dinov2-small) [AI at Meta](https://savrn.com/model-publishers/facebook) Vision Transformer (ViT) model trained using the DINOv2 method. It was introduced in the paper DINOv2: Learning Robust Visual Features without Supervision by Oquab et al. and first released in this repository. Disclaimer: The team releasing DINOv2 did not write a model card for this model so this model card has been written by the Hugging Face team. The Vision Transformer (ViT) is a transformer encoder model (BERT-like) pretrained on a large collection of images in a self-supervised fashion. Images are presented to the model as a sequence of fixed-size patches, which are linearly embedded. One also adds a [CLS] token to the beginning of a sequence to use it for classification tasks. One… Open weights apache-2.0 22M parameters transformers [View model](https://savrn.com/models/dinov2-small) Model · Image feature extraction ### [vit_small_patch14_dinov2.lvd142m](https://savrn.com/models/vit-small-patch14-dinov2-lvd142m) [PyTorch Image Models](https://savrn.com/model-publishers/timm) A Vision Transformer (ViT) image feature model. Pretrained on LVD-142M with self-supervised DINOv2 method. - An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale: https://arxiv.org/abs/2010.11929v2 Explore the dataset and runtime metrics of this model in timm model results. Open weights apache-2.0 22M parameters timm [View model](https://savrn.com/models/vit-small-patch14-dinov2-lvd142m) Model · Image feature extraction ### [clip-vitb-mini-distilled](https://savrn.com/models/clip-vitb-mini-distilled) [AbstractPhila](https://savrn.com/model-publishers/abstractphil) An 8.66M-parameter ViT image encoder (10.0% of a CLIP-B/16 image tower) producing 512-d embeddings compatible with the text tower. The primary checkpoint was distilled on CC12M (10,968,539 images) against the generalized-Procrustes consensus of five CLIP teachers — never against the deployment teacher — and carries a frozen 512×512 rotation that maps its outputs into the deployment frame, where it outperforms the student distilled directly against that teacher on every task gauge, both seeds (full tables below). The rotation is applied by default (config.applyrotation); pass applyrotation=False to getimagefeatures for the raw consensus-frame embedding. Weights are safetensors; the modeling… Open weights apache-2.0 9M parameters transformers [View model](https://savrn.com/models/clip-vitb-mini-distilled) Model · Image feature extraction ### [I-JEPA-Agro](https://savrn.com/models/i-jepa-agro) [Junior R F Junior](https://savrn.com/model-publishers/joserfjunior) Projeto completo de triagem visual de anomalias em folhas de soja usando como extrator de representações e HeraclitusDB como banco de eventos e vetores. 1. lê um manifesto com fazenda, talhão, safra e cultivar; 2. executa o I-JEPA ViT-G/16 localmente; 3. aplica média aos patch tokens e normalização L2; 4. persiste cada imagem e seu embedding euclidiano no HeraclitusDB; 5. cria um artefato local para avaliação offline; 6. treina uma sonda linear; 7. busca imagens visualmente semelhantes; 8. oferece busca por uma API FastAPI. - Python 3.11+; - HeraclitusDB ativo em 127.0.0.1:7474; - vários GB livres para o checkpoint; - RAM suficiente para um modelo com aproximadamente 1B parâmetros; - GPU é… Open weights mit transformers [View model](https://savrn.com/models/i-jepa-agro) Model · Image feature extraction ### [IM-Animation-Motion-Encoder](https://savrn.com/models/im-animation-motion-encoder) [Fynn](https://savrn.com/model-publishers/rbaerk) IM-Animation: An Implicit Motion Representation for Identity-decoupled Character Animation This release contains the final locally retained TiTok-based motion encoder implementation and an exported checkpoint. It provides frame-level motion tokens; it does not include the full animation generator or retargeting network. Code and project videos are available in the linked GitHub repository. The encoder was verified with PyTorch 2.7.1 on CPU using the exported BF16 weights. The weight file is 607,017,208 bytes (approximately 579 MiB). Its SHA256 and provenance are in checkpointinfo.json. frames must be RGB, normalized to [0, 1], with shape [N, 3, 256, 256]. The training preprocessing pads… Open weights apache-2.0 [View model](https://savrn.com/models/im-animation-motion-encoder) ## Google [All models and datasets](https://savrn.com/model-publishers/google) ## Versions - [b4569560a39a](https://savrn.com/models/vit-base-patch16-224-in21k/versions/b4569560a39a) · current 2026-10-24 ## Explore More - [All image feature extraction models](https://savrn.com/models/tasks/image-feature-extraction) - [All models under apache-2.0](https://savrn.com/models/licenses/apache-2-0) - [Model comparisons](https://savrn.com/models/comparisons) - [The model directory](https://savrn.com/models) - [Open model prices by host](https://savrn.com/ai-index/pricing/open-models) ## Source - Repository metadata, read 2026-10-19. - [Google website](https://ai.google.dev/) - [Hugging Face record](https://huggingface.co/google/vit-base-patch16-224-in21k) - [How the hub is built](https://savrn.com/model-hub/methodology)