> ## Documentation Index
> Fetch the complete documentation index at: https://koharu.rs/llms.txt
> Use this file to discover all available pages before exploring further.

# Koharu — AI-Powered Local Manga Translation Tool

> Koharu automates manga translation with ML-powered text detection, OCR, LLM translation, and generative inpainting — all running locally on your device.

Koharu is a local-first desktop application that takes manga from raw scan to finished translation without sending your images to a cloud pipeline. It runs ML models directly on your hardware to detect text regions, read source text with OCR, generate translations through local GGUF models or your chosen hosted provider, and reconstruct the artwork beneath each text block with generative inpainting. Every stage of that pipeline stays inspectable and editable — you can revise OCR output, swap a translation, or repaint a single region without reprocessing the whole project.

<CardGroup cols={2}>
  <Card title="Install Koharu" icon="download" href="/getting-started/install">
    Download and set up Koharu on Windows, macOS, or Linux.
  </Card>

  <Card title="Translate Your First Project" icon="rocket" href="/getting-started/first-project">
    Walk through the complete pipeline from import to export.
  </Card>

  <Card title="Workflow Overview" icon="folder-open" href="/workflow/projects-and-imports">
    Learn how projects, pages, and the editing canvas fit together.
  </Card>

  <Card title="Koharu Agent" icon="robot" href="/agent/setup">
    Use the agent interface to inspect and drive projects programmatically.
  </Card>
</CardGroup>

## What Koharu Does

Koharu organises manga translation into a sequential pipeline. Each stage produces output you can review and edit before the next stage begins. You can run the full pipeline in one pass or open any individual stage to revise its results.

<Steps>
  <Step title="Import and organise">
    Create a project and import pages as individual PNG, JPEG, or WebP images; as CBZ, ZIP, or RAR archives; or as PDFs. Koharu sequences the pages and tracks all project data locally in a single project file.
  </Step>

  <Step title="Detect text regions">
    The Koharu Layout RF-DETR Seg 2XL model scans each page and identifies text regions, speech bubbles, and segmentation masks. Detection runs at page scope or across the full project in one action.
  </Step>

  <Step title="Read source text with OCR">
    An OCR model reads the source text from each detected region. Koharu supports four OCR engines — **PaddleOCR VL 1.6**, **Manga OCR**, **Baberu OCR**, and **Hayai OCR** — so you can choose the engine that performs best on your source material. The OCR output is fully editable before translation begins.
  </Step>

  <Step title="Translate">
    Koharu sends the recognised text to your configured translation engine. Run inference locally with any GGUF-format language model, connect a hosted LLM provider (OpenAI, Gemini, Claude, Grok, DeepSeek, and others), or use a dedicated machine-translation service such as DeepL or Google Cloud Translation.
  </Step>

  <Step title="Inpaint the artwork">
    Before rendered translations appear, Koharu removes the source lettering and reconstructs the artwork behind it. Four inpainting models are available — **FLUX.2 Klein**, **RORem Mixed**, **LaMa**, and **AOT GAN** — ranging from fast structural fill to high-quality generative reconstruction.
  </Step>

  <Step title="Typeset and export">
    Place translated text on the WebGPU-based canvas with automatic fitting, font fallback, vertical CJK support, and right-to-left layout. Export finished pages as flattened PNGs or as layered PSDs for further editing in any image editor.
  </Step>
</Steps>

## Hardware Support

Koharu accelerates ML inference using the GPU backend available on your machine. Keep your graphics driver current; a full CUDA or ROCm SDK installation is not required.

<Tabs>
  <Tab title="CUDA (NVIDIA)">
    CUDA 13.0 is supported on NVIDIA Turing-class or newer GPUs with an R580 or newer driver on **Windows** and **Linux**. Install the [latest NVIDIA driver](https://www.nvidia.com/en-us/drivers/) and verify your GPU against NVIDIA's [CUDA toolkit, driver, and architecture matrix](https://docs.nvidia.com/datacenter/tesla/drivers/cuda-toolkit-driver-and-architecture-matrix.html).
  </Tab>

  <Tab title="ROCm / HIP (AMD)">
    ROCm 10.0 is supported on **Windows** and **Linux** for compatible AMD GPUs. Support depends on your specific GPU, operating system, and driver combination. Check AMD's [ROCm 10.0.0 compatibility matrix](https://rocm.docs.amd.com/en/docs-10.0.0/compatibility/compatibility-matrix.html) and install a compatible [AMD driver](https://www.amd.com/en/support).
  </Tab>

  <Tab title="Metal (Apple Silicon)">
    Metal acceleration is available on **Apple silicon Macs** (M1 and later). No additional setup is required beyond keeping macOS and its graphics drivers up to date.
  </Tab>

  <Tab title="Vulkan">
    Vulkan is available on **Windows** and **Linux** as a cross-vendor alternative to CUDA and ROCm/HIP. Use Vulkan if your GPU is not supported by the CUDA or ROCm backends.
  </Tab>

  <Tab title="CPU Fallback">
    CPU inference is available for all supported workloads on every platform. It requires no GPU and no driver configuration, but runs substantially slower than any GPU-accelerated backend. The editor canvas still requires a current graphics driver because it uses WebGPU regardless of the inference backend.
  </Tab>
</Tabs>

## Privacy and Local-First Design

<Note>
  Koharu runs all vision models and — when you choose a local GGUF model — all LLM inference directly on your machine. Your page images and project data never leave your computer unless you explicitly configure a hosted provider. When you do connect a hosted provider, it receives only the data required for that specific request (for example, recognised text sent for translation), not your raw images or full project.
</Note>

Koharu is available on **Windows**, **macOS (Apple Silicon)**, and **Linux**. Download the latest release from the [GitHub releases page](https://github.com/koharu-rs/koharu/releases/latest), or install with your system package manager — `winget install --id mayocream.koharu` on Windows, `brew install --cask koharu` on macOS.