effect-uai embedding

Turn text and images into vectors. EmbeddingModel is the parallel of LanguageModel — one generic service tag, swap providers at the layer.

Reach for this when the user says any of:

• "I want semantic search / RAG retrieval / a vector index"
• "Embed this text / these documents / this image"
• "Rank these candidates by similarity to a query"
• "Cross-modal retrieval — image query against a text + image corpus"
• "Late-interaction / ColBERT / multivector retrieval"

effect-uai produces vectors. Storage and indexing are userland (or a vector DB).

Install