Search public pages, research tools, and SAVRN solutions.
Open-weight model · Video classification
by Pavel Iakubovskii qubvel-hf/vjepa2-vitl-fpc16-256-ssv2
A frontier video understanding model developed by FAIR, Meta, which extends the pretraining objectives of VJEPA, resulting in state-of-the-art video understanding capabilities, leveraging data and model sizes at scale. The code is released in this repository.
What it takes to serve vjepa2-vitl-fpc16-256-ssv2 (375M 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 | 0.8 GB | 0.9 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 0.4 GB | 0.5 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.2 GB | 0.2 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x 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, read Sep 18, 2026.
By Pavel Iakubovskii, published under mit, revision 99f23ac1d575.
A frontier video understanding model developed by FAIR, Meta, which extends the pretraining objectives of VJEPA, resulting in state-of-the-art video understanding capabilities, leveraging data and model sizes at scale. The code is released in this repository. This is V-JEPA 2 model with video classification head pretrained on Something-Something-V2 dataset. To run V-JEPA 2 model, ensure you have installed the latest transformers
A frontier video understanding model developed by FAIR, Meta, which extends the pretraining objectives of VJEPA, resulting in state-of-the-art video understanding capabilities, leveraging data and model sizes at scale. The code is released in this repository.
To run V-JEPA 2 model, ensure you have installed the latest transformers:
pip install -U git+https://github.com/huggingface/transformers
import torch
import numpy as np
from torchcodec.decoders import VideoDecoder
from transformers import AutoVideoProcessor, AutoModelForVideoClassification
device = "cuda" if torch.cuda.is_available() else "cpu"
# Load model and video preprocessor
hf_repo = "facebook/vjepa2-vitl-fpc16-256-ssv2"
model = AutoModelForVideoClassification.from_pretrained(hf_repo).to(device)
processor = AutoVideoProcessor.from_pretrained(hf_repo)
# To load a video, sample the number of frames according to the model.
video_url = "https://huggingface.co/datasets/nateraw/kinetics-mini/resolve/main/val/bowling/-WH-lxmGJVY_000005_000015.mp4"
vr = VideoDecoder(video_url)
frame_idx = np.arange(0, model.config.frames_per_clip, 8) # you can define more complex sampling strategy
video = vr.get_frames_at(indices=frame_idx).data # frames x channels x height x width
# Preprocess and run inference
inputs = processor(video, return_tensors="pt").to(model.device)
with torch.no_grad():
outputs = model(**inputs)
logits = outputs.logits
print("Top 5 predicted class names:")
top5_indices = logits.topk(5).indices[0]
top5_probs = torch.softmax(logits, dim=-1).topk(5).values[0]
for idx, prob in zip(top5_indices, top5_probs):
text_label = model.config.id2label[idx.item()]
print(f" - {text_label}: {prob:.2f}")
Output:
Top 5 predicted class names:
- Stuffing [something] into [something]: 0.34
- Putting [something] into [something]: 0.25
- Putting [something] onto [something]: 0.04
- Spreading [something] onto [something]: 0.04
- Closing [something]: 0.03
@techreport{assran2025vjepa2,
title={V-JEPA~2: Self-Supervised Video Models Enable Understanding, Prediction and Planning},
author={Assran, Mahmoud and Bardes, Adrien and Fan, David and Garrido, Quentin and Howes, Russell and
Komeili, Mojtaba and Muckley, Matthew and Rizvi, Ammar and Roberts, Claire and Sinha, Koustuv and Zholus, Artem and
Arnaud, Sergio and Gejji, Abha and Martin, Ada and Robert Hogan, Francois and Dugas, Daniel and
Bojanowski, Piotr and Khalidov, Vasil and Labatut, Patrick and Massa, Francisco and Szafraniec, Marc and
Krishnakumar, Kapil and Li, Yong and Ma, Xiaodong and Chandar, Sarath and Meier, Franziska and LeCun, Yann and
Rabbat, Michael and Ballas, Nicolas},
institution={FAIR at Meta},
year={2025}
}
5 files, 1.5 GB in total. The weights are 1 file totalling 1.5 GB in safetensors.
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 1.5 GB | 35aeab96e851 |
| config.json | Configuration | 15.0 KB | — |
| video_preprocessor_config.json | Configuration | 1.5 KB | — |
| README.md | Documentation | 3.6 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
Released by Pavel Iakubovskii through its official repository on Hugging Face. Read the license.
| Precision | Weights in memory |
|---|---|
| As published | 1.5 GB |
| 16-bit | 0.8 GB |
| 8-bit | 0.4 GB |
| 4-bit | 0.2 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
About 0.9 GB at 16-bit and 0.2 GB at 4-bit: the weights (375M parameters) plus a working margin. A long context needs more.
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.
Yes. vjepa2-vitl-fpc16-256-ssv2 is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included.
A frontier video understanding model developed by FAIR, Meta, which extends the pretraining objectives of VJEPA, resulting in state-of-the-art video understanding capabilities, leveraging data and model sizes at scale. The code is released in this repository. This is V-JEPA 2 ViT-L 256 model with video classification head pretrained on Something-Something-V2 dataset. To run V-JEPA 2 model, ensure you have installed the latest transformers
A frontier video understanding model developed by FAIR, Meta, which extends the pretraining objectives of VJEPA, resulting in state-of-the-art video understanding capabilities, leveraging data and model sizes at scale. The code is released in this repository. This is V-JEPA 2 ViT-L 256 model with video classification head pretrained on Diving 48 dataset. To run V-JEPA 2 model, ensure you have installed the latest transformers
VideoMAE model pre-trained on Kinetics-400 for 1600 epochs in a self-supervised way. It was introduced in the paper VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training by Tong et al. and first released in this repository. Disclaimer: The team releasing VideoMAE did not write a model card for this model so this model card has been written by the Hugging Face team. VideoMAE is an extension of Masked Autoencoders (MAE) to video. The architecture of the model is very similar to that of a standard Vision Transformer (ViT), with a decoder on top for predicting pixel values for masked patches. Videos are presented to the model as a sequence of…
A HuggingFace-format conversion of Meta AI's V-JEPA 2.1 ViT-L/16 video encoder and predictor, operating at 384x384 resolution. The weights are Meta's, copied without modification. This repository provides the transformers-compatible packaging plus a documented numerical validation against the original implementation. An equivalent community port already exists (Dev-Jahn/vjepa2.1-vitl-fpc64-384). This repository adds an independently reproduced conversion together with the validation results below. Forward outputs match the existing port to all reported digits. The only structural change is that the fused QKV projection of each attention block is split into separate query / key / value…
A frontier video understanding model developed by FAIR, Meta, which extends the pretraining objectives of VJEPA, resulting in state-of-the-art video understanding capabilities, leveraging data and model sizes at scale. The code is released in this repository. To run V-JEPA 2 model, ensure you have installed the latest transformers: V-JEPA 2 is intended to represent any video (and image) to perform video classification, retrieval, or as a video encoder for VLMs. To load a video, sample the number of frames according to the model. For this model, we use 64. To load an image, simply copy the image to the desired number of frames. For more code examples, please refer to the V-JEPA 2…
Model · Video classification
VideoMAE model pre-trained for 1600 epochs in a self-supervised way and fine-tuned in a supervised way on Kinetics-400. It was introduced in the paper VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training by Tong et al. and first released in this repository. Disclaimer: The team releasing VideoMAE did not write a model card for this model so this model card has been written by the Hugging Face team. VideoMAE is an extension of Masked Autoencoders (MAE) to video. The architecture of the model is very similar to that of a standard Vision Transformer (ViT), with a decoder on top for predicting pixel values for masked patches. Videos are presented to…