Install Koharu¶

Download a release build¶

Download the latest release from the Koharu releases page.

Koharu provides prebuilt binaries for:

Windows
macOS
Linux

If your platform is not covered by a release build, use Build From Source.

What gets installed locally¶

Koharu is a local-first app. In practice, the desktop binary is only part of the installation footprint. The first real run also creates a per-user local data directory for:

runtime libraries used by llama.cpp and GPU backends
downloaded vision and OCR models
optional local translation models you select later

Koharu keeps its own files under a Koharu app-data root and stores model weights separately from the application binary.

First launch expectations¶

On first run, Koharu may:

extract or download runtime libraries required by the local inference stack
download the default vision and OCR models used by detection, segmentation, OCR, inpainting, and font estimation
wait to download local translation LLMs until you actually select them in Settings

This is normal and can take time depending on your connection and hardware.

If you want to prefetch those runtime dependencies ahead of time, run Koharu once with --download. That path initializes the runtime packages and default vision stack, then exits without opening the GUI.

GPU acceleration notes¶

Koharu supports:

CUDA on supported NVIDIA GPUs
Metal on Apple Silicon Macs
Vulkan on Windows and Linux for OCR and LLM inference
CPU fallback on all platforms

Some practical details matter:

detection and inpainting benefit most from CUDA or Metal
Vulkan is mainly the fallback GPU path for OCR and local LLM inference
if Koharu cannot verify that your NVIDIA driver supports CUDA 13.1, it falls back to CPU

For CUDA-capable systems, Koharu bundles and initializes the runtime pieces it needs instead of requiring you to wire every library path by hand.

Note

Keep your NVIDIA driver up to date. Koharu checks for CUDA 13.1 support and falls back to CPU if the driver is too old.

After installation¶

Once Koharu launches successfully, the next decisions are usually:

desktop GUI vs headless mode
local translation model vs remote provider
rendered export vs layered PSD export

See:

Need help?¶

For support, join the Discord server.