Mistral Embed

mistral-embed is a high-performance text embedding model developed by Mistral AI, designed to generate dense vector representations of text for applications such as semantic search, retrieval-augmented generation (RAG), and classification. Supporting input sequences up to 8000 tokens, it produces 1024-dimensional embeddings that capture the semantic meaning of text. The model has demonstrated strong performance on benchmarks like the Massive Text Embedding Benchmark (MTEB), with a retrieval score of 55.26.