# zama-ai/concrete

Concrete: TFHE Compiler that converts python programs into FHE equivalent

Repository: https://github.com/zama-ai/concrete
Canonical: https://ross.abutalabs.com/products/concrete
Language: C++
License: NOASSERTION
License Family: other
Topics: fhe, tfhe, homomorphic-encryption, homomorphic-encryption-library, privacy, gdpr, cryptography, compiler, crypto, encryption, llvm, mlir, python, security, fully-homomorphic-encryption
Last push: 2025-12-19T15:19:04+00:00

## Health v2 (maintenance only)
Score: 56/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 58, release rhythm 28, longevity 100
- inputs: {"age_days": 2162, "days_push": 257, "days_rel": 510, "gap_med": 89, "n_releases_24m": 4}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1574, forks 213 (observed 2026-08-28T04:05:05.758032+00:00)

## What it is
Concrete is an open-source fully homomorphic encryption (FHE) compiler built on TFHE and LLVM that converts Python programs into their FHE equivalents. It lets developers compute on encrypted data without cryptography expertise, with automatic parameter selection and GPU acceleration.

## Use cases
- run computations on encrypted data without decrypting it
- build privacy-preserving services that survive server breaches
- convert existing Python functions into FHE programs
- comply with GDPR by processing user data encrypted
- accelerate homomorphic computations on GPU
- build privacy-preserving machine learning inference (via Concrete ML)

## When to choose
- you need to compute on encrypted data in Python without deep cryptography knowledge
- you want built-in security guarantees and automatic FHE parameter selection
- you need GPU-accelerated homomorphic encryption
- you are building privacy-preserving ML applications on top of Concrete ML

## When to avoid
- you need a general-purpose cryptography toolkit beyond FHE
- your workload cannot tolerate FHE's significant performance overhead
- you need a low-level TFHE library rather than a Python compiler frontend
- you require a permissive OSI license without restrictions (license is BSD-3-Clause-Clear)

## Facets
- artifact type: library
- maturity: active
- function: compiler, cryptography, security, privacy, sdk
- domain: security, privacy, developer-tools, machine-learning
- platform: python, windows, cross-platform
- tags: fhe, tfhe, homomorphic-encryption, llvm, mlir, encrypted-computation, gdpr, cryptography, linux, macos, gpu

## Member repositories
- zama-ai/concrete (main) score 56

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:05.758032+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:57:55.617626+00:00, confidence not recorded.
  - readme: https://github.com/zama-ai/concrete (fetched 2026-08-28T04:05:05.758032+00:00, sha c0c635f090ac)
- Data as of 2026-08-30T08:39:29.467469+00:00.