Which embedding models can I use?

You can choose text-embedding-3-small, text-embedding-3-large, or the legacy ada-002. The 3-small model is the cheapest and fast enough for most retrieval testing, while 3-large gives slightly higher accuracy at higher cost.

What is cosine similarity?

Cosine similarity measures the angle between two embedding vectors, ranging from -1 to 1, where 1 means identical direction and high semantic similarity. Because OpenAI embeddings are normalized, cosine similarity is the standard way to rank candidates by relevance to a query.

Yes. Your key is used only to make the embedding request directly from your browser to OpenAI's API and is never sent to any Gera server or stored anywhere. You can revoke it at any time from your OpenAI dashboard.

Why test retrieval this way instead of a vector database?

For small candidate sets, embedding and ranking in the browser is the fastest way to sanity-check whether semantic search will surface the right documents for your queries. It lets you iterate on phrasing and chunking before investing in a full vector store.

Embedding Nearest Neighbor Finder (BYO-key)

Test semantic search with real embeddings, no infrastructure

Before you build a vector database, you often just want to know: will embeddings actually rank the right document at the top for my queries? This tool lets you paste a query and a list of candidates, embeds them all with your own OpenAI key in a single batched request, and ranks the candidates by cosine similarity — the same operation a vector store performs, but with zero setup.

How it works

You provide your OpenAI API key, a query, and one candidate per line. The tool sends all the texts to the embeddings endpoint in one request, retrieves the vectors, and computes cosine similarity between the query vector and each candidate vector. Because OpenAI embeddings are normalized, cosine similarity is a clean relevance measure: scores near 1 mean strong semantic overlap. Candidates are sorted highest-first so the best match is at the top.

Tips and notes

Your key never leaves your browser except to call OpenAI directly, and it is not stored. Start with text-embedding-3-small — it is inexpensive and accurate enough to validate most retrieval ideas; move to 3-large only if you need the extra precision. This is a great way to iterate on chunking and query phrasing: if the right candidate is not ranking first, the fix is usually better chunk boundaries or a clearer query, not a bigger model. For production, layer a reranker on top of the top results to push relevance even higher.