google/heir

A compiler for homomorphic encryption

What it solves

HEIR provides a universal compiler toolchain for homomorphic encryption (HE), simplifying the process of writing and deploying encrypted computations. It removes the need for developers to manually manage the complex low-level details of different HE libraries and schemes.

How it works

Built on the MLIR (Multi-Level Intermediate Representation) framework, HEIR acts as an intermediate layer that translates high-level encrypted operations into optimized code for specific backends. It supports multiple homomorphic encryption schemes (such as BGV, BFV, CKKS, and CGGI) and integrates with various backend libraries like OpenFHE, Lattigo, tfhe-rs, and Jaxite.

Who it’s for

It is designed for developers and researchers working with homomorphic encryption who want to target multiple HE libraries without rewriting their code for each one.

Highlights

  • MLIR-based toolchain for HE compilers.
  • Support for multiple encryption schemes including BGV, BFV, CKKS, and CGGI.
  • Compatible with several backend libraries including OpenFHE and Lattigo.
  • Provides a Python package (heir_py) for easier integration.

Related

  • Dispatch
    Google HEIR: Enabling Private AI Inference via Homomorphic Encryption

    Google has released HEIR, an open-source compiler that allows pre-trained AI models to operate on encrypted data using homomorphic encryption, aiming to eliminate the trade-off between data privacy and cloud-based AI utility.

  • Project
    iree-org/iree

    IREE is an MLIR-based compiler and runtime that lowers machine learning models to a unified IR for efficient deployment across data centers, mobile, and edge devices.

  • Project
    buddy-compiler/buddy-mlir

    An MLIR-based compiler framework for co-designing DSLs and DSAs, enabling the optimization and deployment of AI models like DeepSeek R1 and Whisper across various hardware targets.

  • Project
    sophgo/tpu-mlir

    An MLIR-based machine-learning compiler that converts pre-trained neural networks from frameworks like PyTorch and ONNX into optimized bmodel files for TPU execution.

  • Project
    google/jsir

    An MLIR-based JavaScript analysis tool that enables high-level source-to-source transformation and low-level dataflow analysis for decompilation and deobfuscation.