Embedding Similarity
Compare two existing numeric vectors with cosine similarity, dot product, Euclidean distance, and Manhattan distance.
Runs entirely in your browser. Nothing you type is uploaded.
About this tool
Paste two existing vectors in bracketed, comma-separated, or whitespace-separated form to calculate four comparison metrics locally. This tool does not generate embeddings; both vectors must already exist and have equal dimensions.
Questions & answers
- Does this tool create embeddings from text?
- No. It only compares numeric vectors that you already have and never calls an embedding API.
- Why can cosine similarity be undefined?
- A zero vector has no direction and a magnitude of zero, so cosine similarity cannot be defined for it.
- Which vector formats are accepted?
- Use values such as [0.1, 0.2], 0.1, 0.2, or numbers separated by spaces and newlines.