Meta publishes MetaEncoder-30B, a gated Apache 2.0 multimodal encoder for decision tasks

Meta has published MetaEncoder-30B on Hugging Face as facebook/meta-encoder. The model card describes it as a multimodal bi-encoder, a "Multimodal System One Encoder", in which both the task and each candidate are written in natural language, with text, images, and video usable on either side. The card says it was built by contrastively fine-tuning Muse-Glimmer-30B and is released under Apache 2.0, inherited from the base model, with use also subject to the base model's Usage Policy; the files can be downloaded only after accepting access conditions on Hugging Face. The accompanying paper was submitted to arXiv on October 8, 2026; the model card does not give a release date. USASI has not tested MetaEncoder-30B.12

USASI. (2026, October 13). Meta publishes MetaEncoder-30B, a gated Apache 2.0 multimodal encoder for decision tasks. United States of America Superintelligence. https://unitedstatesofamericasuperintelligence.com/news/meta-releases-metaencoder-30b/

@misc{usasi_news_meta_releases_metaencoder_30b,
  author = {{USASI}},
  title = {Meta publishes MetaEncoder-30B, a gated Apache 2.0 multimodal encoder for decision tasks},
  year = {2026},
  month = oct,
  howpublished = {\url{https://unitedstatesofamericasuperintelligence.com/news/meta-releases-metaencoder-30b/}},
  note = {United States of America Superintelligence. Last updated 2026-10-13}
}