Nomic Embed is a family of open text embedding models from Nomic AI, released with its weights, training code and training data so anyone can reproduce it.
In plain words
Many embedding models share weights but keep their training data private. The first version, nomic-embed-text-v1, came with everything needed to rebuild it, under the Apache 2.0 licence. It handles texts up to 8,192 tokens. Version 1.5 lets you shorten its 768-number vectors to as few as 64 numbers to save storage, with a small loss in quality. Each text needs a task prefix such as search_query or search_document.
A quick example
In a RAG app, you would store each document as "search_document: " plus its text, and embed each user question as "search_query: " plus the question, then match them by vector similarity.
Where you'll meet it
- Retrieval for RAG and semantic search.
- Local embedding with fully open models.
Related
Sources
- PaperNomic Embed: Training a Reproducible Long Context Text Embedder, Nussbaum et al. (arXiv) · read 29 Sept 2026
- Docsnomic-ai/nomic-embed-text-v1.5, Nomic AI (Hugging Face) · read 29 Sept 2026
Checked by WhatIsAI against the sources above · 29 Sept 2026. Spotted an error? Report it.