OpenMined/TenSEAL

Health v2 · maintenance only

Adoption not part of the score

1033 stars · 174 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

TenSEAL is a Python library (with a C++ core) for performing homomorphic encryption operations on tensors, built on top of Microsoft SEAL. It supports BFV and CKKS schemes for encrypting vectors of integers and real numbers and computing on them without decryption.

Use cases

• perform arithmetic on encrypted vectors without decrypting them
• encrypt machine learning model inputs with homomorphic encryption
• compute dot products and matrix multiplication on encrypted data
• run privacy-preserving inference on sensitive data
• use CKKS encryption for real-number tensors in Python
• access the full Microsoft SEAL API from Python

When to choose

• you need to compute on encrypted numeric data while keeping it confidential
• you want a Python-friendly API over Microsoft SEAL for encrypted tensor math
• you are building privacy-preserving machine learning or secure aggregation pipelines

When to avoid

• you need general-purpose encryption like TLS or file encryption rather than computation on ciphertexts
• performance is critical, as homomorphic encryption carries heavy computational overhead
• you need non-numeric data types or schemes beyond BFV/CKKS

Facets

library · maturity active

cryptography machine-learning math serialization machine-learning privacy deep-learning python cpp windows cross-platform homomorphic-encryption bfv ckks microsoft-seal encrypted-tensors privacy-preserving-computation cryptography linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="OpenMined/TenSEAL")