Get and build the SDK

Use a prebuilt SDK when its platform, backend and API version fit your application. Build from source when you need newer interfaces, a different toolchain or custom backend options. This section covers the native libraries first, then packaging them for an application, Python environment or JVM.

Prebuilt SDKs

The current GitHub release is v1.2.4. The files below contain the native SDK for Windows, Linux, Android, Apple and HarmonyOS. Python uses 1.2.4.post3, including a Windows x64 wheel. The Android AAR remains 1.2.4.post1. Both contain native SDK 1.2.4; their installation links are listed below.

The CUDA/Ubuntu label above is the release asset name; check the library's actual dependencies on the deployment machine. The HarmonyOS archive includes an arm64 native SDK and a staged HAR project with its Node-API library; see HarmonyOS integration to import it. For Windows x64 CPU, use the prebuilt C/C++ SDK, install the Python package from PyPI, or build from source. CoreML and standalone desktop Java have their own build guides below.

Choose the library for the application process, including its architecture and C runtime. A 64-bit device can still run a 32-bit application. Keep each archive’s headers and libraries together, and check the platform guide before linking static frameworks or GPU/NPU libraries.

Apple SDK packaging

Download inspireface-apple-1.2.4.zip for the CPU SDK. One package covers macOS Intel / Apple Silicon, iOS arm64 devices and arm64 / x86_64 simulators. It includes the Objective-C and Swift frameworks used in the examples, along with C/C++ headers and libraries.

InspireFace.framework contains the C and Objective-C interfaces. InspireFaceSwift.framework adds Swift configuration types and scoped buffer helpers. Keep both frameworks from the same build. The package also includes merged platform frameworks, native SDK directories and metadata for checking architectures and deployment targets.

Use the iOS build guide for device and simulator packages, and the macOS build guide for Intel, Apple Silicon and universal packages. iOS frameworks are static; macOS frameworks are dynamic. This distinction controls Xcode's embedding settings. The new iOS framework already incorporates its inference dependency, so it does not need an additional MNN.framework in the application target.

The packaged macOS libraries target 14.0 on arm64 and 15.0 on x86_64. iOS targets 11.0 on devices, 14.0 on arm64 simulators and 11.0 on x86_64 simulators. CoreML packages can be built locally and require a compatible CoreML resource pack; see iOS and macOS for custom builds.

Python and Android packages

The Python package on PyPI is 1.2.4.post3 and includes the 1.2.4 CPU runtime. Install it with:

python -m pip install inspireface opencv-python

To upgrade an existing installation, run python -m pip install --upgrade inspireface. The release provides these py3-none wheels; select the architecture of the Python process:

Windows requires 64-bit Python and the Visual C++ x64 Redistributable. The wheel includes libInspireFace.dll for CPU inference; no local compiler is needed. See Windows integration for local model loading and DLL troubleshooting.

The PyPI file list contains the downloads and checksums. For a custom native library or wheel, see Python packaging and library replacement.

For Android, use the complete 1.2.4.post1 AAR: com.github.HyperInspire:inspireface-android-sdk:v1.2.4.post1, including the leading v. It contains the Android wrappers, complete Java API, models and R8 rules. It supports arm64-v8a, armeabi-v7a and x86_64, with a minimum Android API of 24 and one libInspireFace.so per ABI. Native version queries report 1.2.4, with C API level 2.

The Android integration page includes the repository and Gradle configuration. For local builds, Android builds explains how to produce the JAR and native libraries and add them to an app. Those files are already included when using the AAR.

Java SDK

The Java binding exposes the C API through JNI and runs without Android. Build from the Develop source containing command/build_java.sh. The installed build/java-sdk/install/Java/ directory includes a Java 8-compatible inspireface.jar, native libraries for the target, Java sources and an example.

The JAR is shared across targets; select native libraries for the running JVM’s OS and architecture and keep them paired with the JAR. Integration currently uses this local JAR. See Java packaging for the build and Java integration for a complete program.

Download the model separately

Native SDKs, JVM packages and Python wheels need a resource pack at runtime. The Android 1.2.4.post1 AAR includes Pikachu and Megatron; follow Android initialization to use them, or supply an external pack when changing models. Download packs from the model release: Pikachu and Megatron for CPU, Megatron_TRT for TensorRT, or the Gundam file for the target Rockchip SoC. See model selection and loading for the full mapping.

Choose a build guide