Model Tracker
DINOv3
DINOv3 is an AI model made by Meta. Meta released it on 14 August 2025. It is generally available. Its weights are open. Its licence is DINOv3 License.
- Released
- 14 August 2025 — ai.meta.com, read 16 October 2026
- Status
- Generally available — ai.meta.com, read 16 October 2026
- Takes
- Image — github.com, read 16 October 2026
- Returns
- Not confirmed: The model card describes outputs such as class, patch, and register tokens, which do not map to the register’s output categories (text, image, audio, video, code).
- Open weights
- Yes — github.com, read 16 October 2026
- Licence
- DINOv3 License — github.com, read 16 October 2026
- Context window
- Not confirmed: Meta’s model card and documentation do not print a context-window figure for DINOv3.
- Longest answer
- Not confirmed: Meta’s model card and documentation do not print a standard maximum-output token limit for DINOv3.
- Knowledge cutoff
- Not confirmed: Meta’s model card and documentation do not state a knowledge-cutoff date for DINOv3.
- API price, input
- Not confirmed: The reviewed Meta materials describe released model weights and do not print a standard API input price for DINOv3.
- API price, output
- Not confirmed: The reviewed Meta materials describe released model weights and do not print a standard API output price for DINOv3.
- Where
- Downloadable weights — github.com, read 16 October 2026
- Parameters
- Not confirmed: DINOv3 is a family of 12 models with different parameter counts, not one model with a single parameter figure.
- Model card
- https://github.com/facebookresearch/dinov3/blob/main/MODEL_CARD.md — github.com, read 16 October 2026
Questions
When was DINOv3 released?
Meta released DINOv3 on 14 August 2025.
Is DINOv3 still available?
Yes: it is generally available, as Meta states it.
Are the weights of DINOv3 open?
Yes: Meta publishes the weights of DINOv3 for download.
What is the licence of DINOv3?
The licence of DINOv3 is DINOv3 License, as Meta names it.
More from Meta
- Audiobox
- AudioGen
- Chameleon
- CM3leon
- Code Llama
- DINOv2
- Emu
- Emu Edit
- Emu Video
- EnCodec
- I-JEPA
- ImageBind
- LLaMA
- Llama 2
- Llama 3
- Llama 3.1
- Llama 3.2
- Llama 3.3
- Llama 4 Maverick
- Llama 4 Scout
- M2M-100
- Make-A-Video
- Movie Gen
- Multi-Band Diffusion
Sources
- DINOv3: Self-supervised learning for vision at unprecedented scale (read 16 October 2026)
- dinov3/MODEL_CARD.md (read 16 October 2026)
- dinov3/LICENSE.md (read 16 October 2026)
- facebookresearch/dinov3 (read 16 October 2026)
- DINOv3 (read 16 October 2026)