Skip to content

strands.storage.embeddings.embedder

Provider-agnostic embedder interface for vector search strategies.

Embedder

Section titled “Embedder”
class Embedder(ABC)

Defined in: src/strands/storage/embeddings/embedder.py:6

Produces embedding vectors from text for vector search strategies.

Provider-agnostic: subclass it to plug in any model, or use a vended implementation such as BedrockEmbedder.

embed

Section titled “embed”
@abstractmethod
async def embed(text: str) -> list[float]

Defined in: src/strands/storage/embeddings/embedder.py:14

Embed one text.

The vector length must match the dimension of the index it is stored in.

Arguments:

  • text - The text to embed.

Returns:

The embedding vector.