Open-source project
kmonkeyhead/MORT avatar
kmonkeyhead/MORT

MORT: Real-time OCR game translation for Windows

MORT 번역기 프로젝트 - Real-time game translator with OCR

1,811 stars89 forksC#MIT

At a glance

What is it?
MORT is a Windows OCR and translation tool for reading game dialogue from selected screen regions. It supports several OCR engines and translation services, but it cannot capture fullscreen games and depends on the chosen engine, language settings, and region selection.
Who is it for?
Adopt MORT when you play Windows games in windowed or borderless windowed mode and need translated dialogue from a selected screen area. Do not choose it for fullscreen titles, 64-bit systems whose CPU lacks AVX2, or setups that require an undocumented installation path.
Can I use it commercially?
Yes. MIT is a permissive licence: you can use, modify and sell software built on it, as long as you keep its copyright and licence notices.
Is it still maintained?
Yes. The repository last received commits 42 days ago.
What is it written in?
Mainly C#, according to GitHub's language statistics.

Answers come from the project's GitHub data, last synced on October 22, 2026, and from our analysis. They are not legal advice.

Editorial analysis

MORT's role in translating game dialogue

MORT addresses a specific problem: game dialogue is often displayed as pixels rather than selectable text, so ordinary text translation tools cannot read it. MORT captures a region of the game screen, applies optical character recognition, and sends the recognized text to a translation database or machine translation service. The result is intended to appear as a real-time translation while the game remains visible.

The project is for Windows players who need to understand dialogue in games, particularly when the game supports windowed or borderless windowed display. It is not a general-purpose document translator, and the README does not describe support for reading text outside the screen region selected by the user. The supported extraction path is also constrained by the OCR engine and the language selected for that region.

The README identifies English and Japanese translation and extraction as defaults. It also lists language patching through a database and OCR-based extraction from multiple areas. This makes MORT relevant to games where dialogue is localized through images or where the player cannot easily access the original text. The project does not claim automatic recognition of every game interface, font, or language.

The OCR and translation path through selected regions

The documented flow begins with a selected OCR area. After configuring quick settings, the user starts translation from the remote control, selects the screen region where dialogue appears, applies that selection in the main MORT window, and then presses Translate. Based on this workflow, MORT watches the selected area, extracts text with OCR, and passes the resulting text to a configured translation backend.

MORT supports TesseractOCR, Windows OCR, Google Cloud Vision OCR, Snipping Tool OCR, and Easy OCR. The selected engine can affect recognition of small fonts, stylized lettering, transparency, motion, or low-contrast text. The README does not publish a comparison of accuracy, latency, or resource consumption between these engines, so it does not support a claim that one is consistently better for every game.

Translation backends include Naver Papago, Google Web, Google Sheet, ezTrans, and DeepL. MORT can also use a database for language patches. For a custom HTTP integration, the documented POST request contains `name`, `text`, `target`, and `source`. The expected response contains `result`, `errorCode`, and `errorMessage`. An external service must conform to this contract for the documented Custom API path to return a usable translation.

Downloading MORT and performing the first translation

MORT does not provide a package-manager installation command in the README. It directs users to the project's Naver blog for the latest version download and release notes. The documented runtime requirements are Windows 10 or higher, a 64-bit operating system for the 64-bit application, .NET 9 or higher, and Visual Studio 2022 Visual C++ x64 runtime. The README links to the Microsoft .NET desktop runtime installer and the Visual C++ redistributable.

The linked runtime download is shown here exactly as listed in the README:

text
https://dotnet.microsoft.com/ko-kr/download/dotnet/thank-you/runtime-desktop-9.0.4-windows-x64-installer

The Visual C++ runtime is identified in the README as `vcredist_x64.exe`, with its download link shown separately. After obtaining a build, open the Preferences tab and set the OCR language to match the language used by the game. On the remote control, select Search, mark the region where dialogue appears, return to the MORT main form, and click Apply. Press Translate on the remote control to start translation. The expected result is text recognized inside the selected region and translated through the configured backend.

The README does not document a command-line installer, portable package layout, or unattended deployment process. It also does not explain whether cloud services require credentials in this article. Users should follow the linked release notes and the in-application preferences rather than infer an unsupported installation method.

Adding result languages and using a Custom API

MORT lets users add a Google Translator language code in `UserData/UserTransCode.txt`. The README presents the file as a two-column list containing a code and a name, with Italian and `it` given as the example. It also links to Google Cloud's language documentation for the available codes. This is a local customization point for output languages rather than a way to install additional OCR models.

The Custom API option is intended for an HTTP-based translation service. The README directs users to choose Custom API under Translation Type and configure the address through Advanced Settings. Its request rule specifies four string fields: `name`, recognized OCR text in `text`, the translation result language in `target`, and the OCR language in `source`.

The response rule requires `result` for translated text, `errorCode` for an error code, and `errorMessage` for an error message. The README links to separate example projects for LibreTranslate and NLLB. Those examples demonstrate possible service integrations, but the README does not say that either service is bundled with MORT. Authentication, retries, rate limits, and behavior after malformed responses are not documented, so an operator cannot rely on unstated network handling.

Why MORT fails outside its documented capture boundary

The clearest limitation is display mode. The README explicitly says that fullscreen games cannot be used and recommends windowed mode or borderless windowed mode instead. MORT depends on selecting a visible screen region, while the FAQ excludes fullscreen games. This is a compatibility boundary, not a setting that the README says can be enabled.

Architecture also affects installation. The primary system requirements call for a 64-bit OS, while the FAQ provides a separate 32-bit version for users who cannot use the 64-bit build. On 64-bit Windows, error `0x8007045A` is associated with a CPU that does not support AVX2. The documented workaround is to use the 32-bit version. An installation that satisfies the Windows version requirement can still fail to start when the selected build requires CPU features the machine does not support.

The README lists Tesseract OCR 5.2.0 and NHocr 0.21 as development dependencies. It also warns that building the project does not necessarily produce a runnable application because essential runtime files are required. Its build instructions involve extracting the latest build files into the release folder and refer to separate components for modifying `MORT_CORE.DLL` and `nhocr.DLL`. This makes source-level setup more involved than downloading an ordinary application package.

Maintenance, upgrades, and MIT licensing

MORT is not archived, and its last push was on 2026-09-11. The listed releases include version 1.320 on 2026-09-11, version 1.319 on 2026-08-23, and version 1.318 on 2026-08-15. These dates show recent repository activity, but the README does not promise a release schedule, support term, or compatibility policy.

The repository is licensed under the MIT License. That identifies the license associated with the project, but this article does not interpret the license or provide legal advice. The top-level layout includes separate directories for cloud vision, Google translation, the updater, pipe communication, documentation, packages, and tools. This indicates several components around the main MORT application, although the README does not provide a formal matrix showing how each one changes between releases.

Upgrade cost may involve more than replacing the executable. Users must account for OCR engine changes, .NET runtime requirements, language-code configuration, database patches, and custom API response fields. Developers are directed to separate repositories for `MORT_CORE_DLL` and `nhocr.DLL`. The linked user manual and Trello development plan may provide additional detail, but their contents are not reproduced in the repository README available here.

Editorial conclusion

Adopt MORT when you play Windows games in windowed or borderless windowed mode and need translated dialogue from a selected screen area. Do not choose it for fullscreen titles, 64-bit systems whose CPU lacks AVX2, or setups that require an undocumented installation path. Verify the downloaded release, OCR language, selected translation service, region selection, and fullscreen compatibility first. MORT's documented boundary is its need for Windows, .NET 9 or higher, 64-bit hardware for the 64-bit version, and an OCR-readable dialogue area.

Frequently asked questions

Can MORT translate a game running in fullscreen mode?

No. The README says that fullscreen games are unsupported and recommends windowed mode or borderless windowed mode instead.

What are the system requirements for MORT?

The README lists Windows 10 or higher, a 64-bit operating system, .NET 9 or higher, and the Visual Studio 2022 Visual C++ x64 runtime. It separately mentions a 32-bit version for systems that cannot use the 64-bit build.

Which OCR and translation services does MORT support?

The README lists TesseractOCR, Windows OCR, Google Cloud Vision OCR, Snipping Tool OCR, and Easy OCR for OCR. It lists Naver Papago, Google Web, Google Sheet, ezTrans, DeepL, and a Custom API for translation.

Official sources

  1. kmonkeyhead/MORT on GitHub
  2. License: MIT
  3. Project website
  4. README
  5. Releases
Add this badge to your README

If you maintain this project, the badge below links readers to this analysis and shows its maintenance status from the daily GitHub snapshot. Paste the markdown into your README; add ?metric=license or ?metric=stars to the image URL for a different field.

Add this badge to your README

markdown
[![Hysen Labs](https://hysenlabs.com/badge/kmonkeyhead-mort.svg)](https://hysenlabs.com/projects/kmonkeyhead-mort)

Was this useful?