购物车 (
Forgotten the filename? Find it again with Gety, a local AI file search tool now available
数码荔枝编辑部
It’s new-release Friday again!
Do you often find yourself searching your entire hard drive for a file? You clearly remember saving it, but simply cannot recall the file name. The system search is not only slow, but also fails to find text in images.
Gety, now available this week, is a local AI file search tool designed to solve this kind of problem, with support for both Windows and Mac.
Gety is the latest creation from the team behind the long-established productivity tool Listary ——
If you have used Listary, you will certainly remember its ability to launch software quickly and search for files by name. This time, the development team has extended its search capabilities from ‘file names’ to ‘file contents’, allowing you to find files by their contents even if you have forgotten their names.
Lightning-fast full-text search
When Gety is launched for the first time, it quickly indexes the entire drive. You can start using it within a few minutes, and every subsequent search returns results within milliseconds.
Simply type in your query and the search results will refresh instantly, without having to press Enter and wait. Compared with system searches that can spin for dozens of seconds, the difference in experience is huge.
It supports PDF, Word, Excel, PPT, WPS, TXT, Markdown, HTML and images (via OCR), covering most commonly used file types.
File changes are detected automatically and the index is updated incrementally, so there is no need to rebuild it manually. Search results are sorted by relevance, with the most relevant content shown first.
Search scanned documents too
Traditional search tools are helpless when it comes to scanned PDFs, screenshots and photos, because the text in these files is an image rather than text characters.
Gety has a self-developed local OCR engine built in. It works as soon as the software is installed, automatically extracting text from images during indexing so it can be searched just like ordinary documents. For scenarios involving scanned documents or text in images, it also reduces the hassle of searching manually.
▲ Images in documents can also be processed with OCR
It can recognise webpage screenshots, chat records, scanned contracts and even tilted or distorted documents. Mixed Chinese and English text is no problem either.
Recognition accuracy is lower for handwriting and extremely low-resolution images, but it is already sufficient for everyday office work. New screenshots and scanned documents are also detected and extracted automatically.
Preview rendered from the original document
Many search tools only extract the plain text from a document for preview, removing the original layout and tables so that it looks different from the source document.
Gety’s previews are rendered from the original document: PDFs retain their PDF layout, Word files retain their styles, and Excel files display their tables…
Open a result to highlight the matching passages. The match count is shown in the top right, and you can click the up and down arrows to jump between each match.
▲ Markdown can be rendered or displayed as plain text
You can even select and copy text recognised by OCR directly in the preview, without opening another application, saving you the time spent repeatedly checking the original.
Fuzzy semantic search
Sometimes we only remember the general meaning and cannot recall the original file name or its actual contents. With Gety, you can still find it easily!
For example, searching for ‘agreement’ can also find the ‘distribution contract’ that Lizhi actually wants.
Gety’s semantic search works through a locally embedded model, without calling any external API. It is enabled by default and works alongside keyword matching.
Results found semantically are marked with an ‘AI semantic match’ label, making them easy to distinguish and helping you quickly recover those documents whose ‘general meaning is all you remember’.
A local source of truth for AI
Although AI tools can read local files, asking them to search through an entire computer from scratch for relevant content means scanning the whole drive for every task. This is slow, consumes a large number of additional tokens, and requires you to grant the AI Full Disk Access, which carries considerable risk.
Gety acts as a local search engine. AI can call it via MCP, API or CLI to retrieve the most relevant content snippets within milliseconds and feed them directly to the AI as context, without making the AI search through the entire hard drive itself.
By default, it only returns the relevant snippets you have permitted, so you do not need to hand over all the data on your hard drive. Combined with a local model, it can remain entirely offline, keeping privacy under control. For users concerned about the security of local AI search, this on-demand approach to providing content makes permissions easier to manage.
Connectors can also bring in cloud data such as browser history, Notion and Feishu, so AI searches are not limited to the local hard drive.
In conclusion
Gety turns ‘searching through your computer for a file’ into ‘typing and getting results instantly’. From searching and previewing to copying or dragging and sending, the whole process generally takes around 10 seconds!
For a frequent task such as ‘finding files’, it can save a great deal of time over the course of a day.
Gety is now available from the lizhi.shop software store, where you can give it a try. If you want to unlock more features, you can learn more about the full capabilities of the Pro version and then choose the version that best suits your file management and AI workflow needs.
If you frequently need to find files, Gety is well worth adding to your daily workflow as a local AI file search tool, bringing faster and more accurate file searching for both you and your AI.
这里空空如也，暂时~