Merck/Sapiens

Sapiens is a human antibody language model based on BERT.

62
/ 100
Established

Enables humanness scoring, sequence infilling with masked residue prediction, and mutation suggestion for antibody design through separate VH/VL models hosted on HuggingFace. Built on masked language modeling architecture, it generates per-position residue probability scores and sequence embeddings that capture immunological properties learned from natural antibody repertoires. Integrates with HuggingFace transformers ecosystem and provides mutation suggestion capabilities for both framework and CDR regions without requiring mask tokens.

Available on PyPI.

Maintenance 10 / 25
Adoption 15 / 25
Maturity 18 / 25
Community 19 / 25

How are scores calculated?

Stars

72

Forks

20

Language

Jupyter Notebook

License

MIT

Last pushed

Jan 22, 2026

Monthly downloads

564

Commits (30d)

0

Dependencies

3

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/embeddings/Merck/Sapiens"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.