anymize Embeddings icon

anymize Embeddings

Create embeddings with anymize for vector stores

anymize Embeddings

AI-generated

Summary

Creates embeddings for text supplied by a connected vector store, sending requests to the anymize embeddings API, which processes text with Voyage AI. Configure the model and optional batch size, output dimensions, retry limit, and request timeout; dimensions should match the vector store index.

Inputs

  • anymizeApi credentials (required) — API key used to authenticate requests.
  • Model — Embedding model name; defaults to voyage-4-lite, which the UI identifies as the only available model.
  • Batch Size — Maximum documents per request; defaults to 512 and is capped at 1,000.
  • Dimensions — Requested embedding size: 256, 512, 1,024 (default), or 2,048. Match this to the vector store index.
  • Max Retries — Maximum retry count; defaults to 2.
  • Timeout — Request timeout in milliseconds; defaults to 60,000.

Output shape

An n8n AI Embeddings connection providing an embeddings provider to a connected vector store.

The provider embeds documents in batches and embeds queries individually. Each API response’s vectors are ordered by their returned index. The node does not emit regular workflow items; its connection trace reports the input and embedding count and dimensions.

Links

Discussion