Overview

This model was trained with multilingual knowledge distillation so that sentences with the same meaning in different languages get similar vectors.

Highlights

  • 278M parameters
  • 50+ languages
  • 768-dimensional embeddings
  • Apache 2.0 license

Typical uses

Cross-language similarity, clustering and lightweight search.

Things to consider

  • Lower retrieval quality than modern embedding models.
  • Download sizes refer to the default quantized versions in the Ollama library; full-precision weights are larger.

Specifications are taken from the model card on Hugging Face, the Ollama library and the provider’s announcement. Benchmark results depend on evaluation settings; see how AI benchmarks work.