Skip to content
RAG Repo

Cohere Wikipedia 22-12 Embeddings (per-language)

This dataset encodes Wikipedia using the Cohere multilingual-22-12 embedding model, converting each passage into an embedding, a numeric vector that captures its meaning, computed on the article's title plus its text. It is split by language, so you can pull down just the languages you need rather than the whole collection.

Coverage spans many languages, including Arabic, German, English, Spanish, French, Hindi, Italian, Japanese, Korean, Simple English, and Chinese. Because the passages are already embedded, you can load a language into a vector database and run semantic search, matching results by meaning rather than exact keywords, without building an embedding pipeline yourself.

The files can be streamed instead of downloaded in full, which is helpful given the size. The embeddings are released under Apache 2.0, while the underlying Wikipedia text stays under CC BY-SA, and whether embeddings of ShareAlike text count as a derivative work remains unsettled.

wikipediaembeddingsmultilingualsemantic-searchrag-readystreaming

Related sources