AxForge
Sign in Get started

Model reference · open weights

dinov3-vith16plus-pretrain-lvd1689m

Embeddings facebook Image embed 1 build Its own licence terms 65k dl/mo

dinov3-vith16plus-pretrain-lvd1689m is an open-weight embedding model from Meta. dinov3-vith16plus-pretrain-lvd1689m (FP32) weighs 1.7 GB; the smallest configuration that runs it is RTX 3060 12 GB.

  • DINOv3-vith16plus-pretrain-lvd1689m is a vision foundation model developed by Meta for image feature extraction.
  • It is a Vision Transformer with 841M parameters that uses a patch size of 16 and was pretrained on the LVD-1689M web dataset.
  • The model is distributed under a custom license.

Summary of the facebook/dinov3-vith16plus-pretrain-lvd1689m model card, 2026-10-01

Hugging Face ↗GitHub ↗Paper ↗

What it is

Released byMeta
Released2025-08-06
Parameters841M
VRAM1.7 GB for the weights

What it runs on

Memory and cards for dinov3-vith16plus-pretrain-lvd1689m (FP32)

1.7 GBweights, file size
1.1 GBruntime overhead
CardRunsMemory
RTX 3060 12 GBfits11.6 GB
RTX 4060 Ti 16 GBfits15.4 GB
RTX 3090 24 GBfits23.4 GB
RTX 4090 24 GBfits23.4 GB
RTX 5090 32 GBfits31.0 GB
L40S 48 GBfits44.0 GB
A100 80 GBfits78.2 GB
H100 80 GBfits78.1 GB
RTX PRO 6000 Blackwell 96 GBfits93.8 GB
DGX Spark (GB10) 128 GB unifiedfits107 GB
H200 141 GBfits138 GB
B200 180 GBfits176 GB

How it works

How embedding models work

Your textsentence / documentEncodermaps meaningVectorlist of numbersAn embedding model turns text into a vector, so similar meanings sit close together, which is the basis of search and RAG.
© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms