DigUp brings content-based search to Mac files with Google DeepMind’s EmbeddingGemma 2

News Editor

2026-10-14 11:03:43

Open-source Mac app DigUp is shifting file search away from filenames and keyword recall toward content-level retrieval. Built by independent developer Abdur Rahim, the tool lets users search with natural descriptions such as "the zebra in the video" or "the part where sleep was discussed," then jump to the matching video frame or podcast timestamp. It also works across photos and PDFs, including opening the page where a specific contract clause appears. DigUp uses EmbeddingGemma 2, a model released by Google DeepMind on Oct. 6 that converts text, images, audio, and video into comparable embeddings. The app creates an index for selected folders in advance and can also match exact items such as invoice numbers and error codes. According to BlockBeats, all indexing and search run locally on the Mac, so private files do not need to be uploaded to the cloud. DigUp is free and open source, has picked up more than 3,000 GitHub stars in less than a week, and currently supports only Apple silicon Macs running macOS 14 or later. First-time setup requires downloading a model of about 865 MB and waiting for indexing to finish.

Open-source app DigUp is reworking file search on the Mac by moving beyond filenames and basic keywords. If a user only remembers that a zebra appeared in a video, or that a podcast episode mentioned sleep, DigUp can search for the content itself and point to the relevant result.

Search by what appears in media, not just by file labels

According to BlockBeats, DigUp can take a query such as "the zebra in the video" and jump to the frame where the zebra appears. A search for "the part where sleep was discussed" can locate the relevant timestamp in a podcast.

The app also searches photos and PDFs. It can even open the page where a specific clause appears in a contract, extending search from metadata into the actual contents of files.

Built on EmbeddingGemma 2 and processed locally on Mac

DigUp uses EmbeddingGemma 2, released by Google DeepMind on Oct. 6. The model turns text, images, audio, and video into numerical representations that can be compared across formats, allowing a text description to match a related image or sound.

The app builds an index for selected folders ahead of time. During search, it can also match exact items such as invoice numbers and error codes. BlockBeats said the full indexing and search workflow runs locally on the Mac, without uploading private files to the cloud.

Free and open source, with Apple silicon support only for now

DigUp is free and open source. In less than a week after release, it has gained more than 3,000 GitHub stars.

At this stage, the app supports only Macs with Apple silicon and requires macOS 14 or later. First-time use also requires downloading a model of about 865 MB and spending time to complete indexing.

This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan.

1000

Disclaimer:

The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.

Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.