Model Tracker
Perception Encoder Audiovisual (PE-AV)
Perception Encoder Audiovisual (PE-AV) is an AI model made by Meta. Meta released it on 16 December 2025. It is generally available. Its weights are open.
In inglese
- Released
- 16 December 2025 — ai.meta.com, read 16 October 2026
- Status
- Generally available — huggingface.co, read 16 October 2026
- Takes
- Text, Audio, Video — huggingface.co, read 16 October 2026
- Returns
- Not confirmed: The model returns embeddings, which are not among the register's allowed output categories (text, image, audio, video, code).
- Open weights
- Yes — huggingface.co, read 16 October 2026
- Licence
- Not confirmed: two readings disagree: Apache 2.0 and Apache-2.0
- Context window
- Not confirmed: The reviewed maker pages do not state a standard context window in tokens for PE-AV.
- Longest answer
- Not confirmed: The reviewed maker pages do not state a standard maximum answer length in tokens; PE-AV is described as an encoder that returns embeddings.
- Knowledge cutoff
- Not confirmed: The reviewed maker pages do not state a knowledge cutoff for PE-AV.
- API price, input
- Not confirmed: The reviewed maker pages do not publish a standard API input price for PE-AV.
- API price, output
- Not confirmed: The reviewed maker pages do not publish a standard API output price for PE-AV.
- Where
- Downloadable weights — huggingface.co, read 16 October 2026
- Parameters
- Not confirmed: PE-AV is released as multiple checkpoints of varying sizes; the reviewed pages do not specify one parameter count for the model family as a whole.
- Model card
- https://huggingface.co/facebook/pe-av-base — huggingface.co, read 16 October 2026
Questions
When was Perception Encoder Audiovisual (PE-AV) released?
Meta released Perception Encoder Audiovisual (PE-AV) on 16 December 2025.
Is Perception Encoder Audiovisual (PE-AV) still available?
Yes: it is generally available, as Meta states it.
Are the weights of Perception Encoder Audiovisual (PE-AV) open?
Yes: Meta publishes the weights of Perception Encoder Audiovisual (PE-AV) for download.
More from Meta
- Audiobox
- AudioGen
- Chameleon
- CM3leon
- Code Llama
- DINOv2
- DINOv3
- 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
Sources
- Introducing SAM Audio: The First Unified Multimodal Model for Audio Separation (read 16 October 2026)
- Perception Models: Powerful Models for Image, Video, and Audio Perception (read 16 October 2026)
- Perception Encoder Audio-Visual (PE-AV) (read 16 October 2026)