SunnyCapturer: a screenshot tool with OCR, translation and pinned images

The capture tool, plus four things around it

SunnyCapturer is described as a simple and beautiful cross-platform screenshot software, and the feature line is what distinguishes it from the dozen other capture utilities on any given desktop. Beyond taking a screenshot, it supports OCR for extracting text from images, image translation, custom stickers, and pinning images to the screen. Those are the four additions that turn a capture tool into something you keep running all day. OCR is offline, which is the detail that matters for a tool that runs on screenshots of whatever is currently on your screen, since an online service would be sending that content to a third party every time you captured something. Image translation works on the captured image rather than on selected text, which is what makes it useful for reading foreign-language text in a screenshot. Sticker and pin features address the two things people actually want to do with a capture: decorate it before sending it, and keep it visible as a reference while working on something else. The primary language of the project is C++, and the release tags move quickly, with v5.3.0 on 2026-04-02, v5.2.0 on 2026-03-09 and v5.1.0 on 2026-02-27.

The documented platforms, including Loongson and ARM64

The README shows the application running on a wide set of platforms, and the detail is worth reading because a screenshot tool is unusually sensitive to the display stack of the operating system it runs on. On Windows there are exhibition images labelled for Windows 11. On Linux the shots cover Debian 12, Ubuntu 20.04 through 24.04, Deepin V23 and OpenKylin 2.0. On macOS the range given is versions 13 through 15. And there is a separate section for Linux Deepin V23 on ARM64 and on Loongson, with images for each. The presence of Loongson is unusual enough to call out, because that is a domestic Chinese instruction set architecture rather than x86 or ARM, and supporting it means the project is built for environments where the usual toolchain assumptions do not hold. The ARM64 build is the other one that matters in practice, since that is the architecture of the machines most likely to be running Linux today. The README points to a separate document, supported_os.md, for details of the supported operating system versions, so the list in the README is a summary rather than the full matrix. The last push to the repository was on 2026-04-02, matching the v5.3.0 release date, and the repository is not archived.

Where to get it, and where the documentation lives

The README does not include an install command, and the project does not expect one, because SunnyCapturer is distributed as a finished application rather than as something you build from source. The official website and downloads and user guide are at sunnycapturer.xmuli.tech, the Microsoft Store listing is linked directly, and there is a separate source code repository at the SunnyCapturer organisation on GitHub, with a separate translations repository where contributors are asked to help. That separation is a bit confusing at first, because the repository you are reading is not the one hosting the code. The top level of this repository is mostly documentation and a static site, with index.html, a _sidebar.md, a CNAME file, a _coverpage.md, a favicon, a logo, a _media directory holding the screenshots, and a zh-cn directory, which is the Chinese version of the same site. The documentation files at the top level are organised as a set of pages, with getting_started.md, development_technology.md, extract_text_ocr.md, image_translation.md, translation.md, troubleshooting.md, supported_os.md, office_site.md, acknowledgement.md, changelog.md, and a history-version-code directory for older versions. That is an unusually complete documentation set for a utility, and it is where the install steps, the supported versions and the feature details actually live.

Feedback, translations and a project run by one person

The community arrangements are described in two short sections. For feedback, the README asks you to check the closed issues first, and if nothing similar exists, to open a new issue so that bugs and suggestions can be tracked and resolved effectively. For translations, there is a dedicated repository under the same organisation, which is the model that works for a project with a user base that reads in more than one language. The contact details point at a QQ group, a GitHub profile and an email address, which tells you something about where the users are. The acknowledgements page and the changelog file are both checked in, and the acknowledgements page existing as a separate document is the mark of a project that credits the people who helped it. The overall picture is of a single-maintainer utility with a Chinese-speaking user base and an international reach, with the website, the store listing and the Discord-style community contact as the visible surface. There is no contributing guide, no code of conduct file and no development setup documentation in this repository, which is consistent with the code living elsewhere.

The licence question, and the version history directory

The licence is the one thing the repository does not settle, and it matters more here than it would for a library. The project listing records no licence identifier at all, and there is no licence file among the top level entries, which run from .github and .gitignore through the documentation files to the zh-cn directory. For an application you download and run on your own machine, an unstated licence is unusual but survivable, since you are not redistributing anything. It becomes a problem the moment you want to package SunnyCapturer into a product, ship it inside a corporate image, or read the source to modify it, and the answer at that point has to come from the maintainer. The history-version-code directory at the top level is also worth a note, since it suggests that older versions of the source are kept in the tree, which is a helpful thing for anyone who wants to compare how the capture code changed, and it is another sign that this repository is a documentation and archive site rather than the active development tree. The presence of a .nojekyll file and a CNAME file confirms the site is served as a static GitHub Pages site.

Where SunnyCapturer sits against the built-in tools

Every desktop now has a capture shortcut built in, and on macOS that shortcut is genuinely good, with a preview, annotations and a share sheet. So the case for a third-party tool has to rest on the four additions rather than on the capture itself, and the additions are not evenly weighted. Offline OCR is the one that changes daily work, because it turns a screenshot of a terminal, an error dialog or a paragraph into selectable text, and the fact that it is offline makes it usable on content you would not send to a service. Image translation is the second, and it overlaps with the OCR in a useful way since the text you extract can be the text you translate. Pinning and stickers are the third and fourth, and they are conveniences rather than differentiators, since a window manager or a screenshot tool with a pin mode covers them. The limit of the comparison is that SunnyCapturer is a single application with a fixed feature set, whereas the built-in tools are part of an operating system that can also be extended. A user who needs a specific workflow, such as capturing a scrolling region with text recognition baked into the annotation step, may find a general purpose tool with an extension mechanism a better fit. What SunnyCapturer offers is one coherent set of features, already built and working, on three operating systems including the ones that rarely get good tooling.

Editorial conclusion

Adopt SunnyCapturer if you take a lot of screenshots and want text extraction, translation and pinned reference images in the same tool on Windows, Linux or macOS, and you are happy with a project whose licence is not declared in the repository. Skip it if you need a screenshot tool with a clear open source licence, since none is stated here, or if you need an audited build, since the documented install route is a signed store package or a download from the project site. Verify first which operating system versions the project lists as supported, whether your Linux distribution and architecture are among them, and what the terms are for redistribution before you package it into anything.

Frequently asked questions

Official sources

1. Issues
2. Project website
3. README
4. Releases
5. XMuli/SunnyCapturer on GitHub