From the model card
Fields below are copied from the tags and counters on the HuggingFace repository nomic-ai/nomic-embed-text-v1.5 at our last fetch. They are set by the uploader, not verified by us; rows with no tag are omitted. How this page is made.
- Publisher (HF namespace)
- nomic-ai
- Pipeline tag
- sentence-similarity
- Library
- Sentence Transformers, Transformers, Transformers.js
- Weight formats
- ONNX, safetensors
- License tag
apache-2.0— read the license file in the repo before relying on it- Language tags
- English (en)
- Papers cited
- arXiv:2402.01613, arXiv:2205.13147
- Downloads (HF counter at last fetch)
- 15,943,256
- Likes (HF counter at last fetch)
- 901
- Model card
- https://huggingface.co/nomic-ai/nomic-embed-text-v1.5
Use cases
- RAG pipeline text embedding with flexible dimension budget
- Semantic search where embedding size can be tuned to vector store cost
- Browser-side embedding inference via transformers.js without a server
- MTEB benchmark comparison against other embedding models
- Building efficient embedding pipelines where 768 dims is over-budget
Pros
- Matryoshka dimensions allow truncating to smaller sizes without significant accuracy loss
- Transformers.js compatibility enables client-side or edge inference
- Apache 2.0 license; ONNX and safetensors supported
- MTEB retrieval scores competitive with larger models
- Custom nomic-BERT architecture trained specifically for retrieval
Cons
- English-only; no cross-lingual capability
- Custom nomic_bert architecture requires custom_code flag — less standard than BERT-based models
- Smaller adoption footprint than sentence-transformers standard models
- Performance at smallest dimensions (64d) degrades on hard retrieval tasks
- Requires trusting third-party custom model code on load
Tags
sentence-transformersonnxsafetensorsnomic_bertfeature-extractionsentence-similaritymtebtransformerstransformers.jscustom_codeenarxiv:2402.01613arxiv:2205.13147license:apache-2.0model-indexeval-resultstext-embeddings-inferenceendpoints_compatibleregion:us