| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| latest.json | 2026-10-08 | 1.2 kB | |
| Semantra_aarch64.app.tar.gz.sig | 2026-10-08 | 408 Bytes | |
| Semantra_aarch64.app.tar.gz | 2026-10-08 | 910.0 MB | |
| Semantra_0.2.2_aarch64.dmg | 2026-10-08 | 919.3 MB | |
| README.md | 2026-10-08 | 2.4 kB | |
| Semantra v0.2.2 source code.tar.gz | 2026-10-08 | 572.4 kB | |
| Semantra v0.2.2 source code.zip | 2026-10-08 | 633.3 kB | |
| Totals: 7 Items | 1.8 GB | 0 | |
(A.I.-generated text follows): Semantra 0.2 is a big rewrite. It now uses Google's EmbeddingGemma 2, which runs entirely on your machine through Apple MLX. You can search text, images, PDF pages, audio and video together, and you can search with text, an image, an audio file or your voice.
Requirements: an Apple Silicon Mac (M1 or later). For Windows, Intel Macs, or the older smaller model, use v0.1.3.
Download
- Semantra_0.2.2_aarch64.dmg: open it and drag Semantra to Applications. The app is signed and notarized. The download is large (~920 MB) because the model weights are bundled.
What's new in 0.2.2
- In-app updates. From this version on, Semantra checks for new releases on launch and offers Install and restart. (Coming from an earlier version, you need to install 0.2.2 by hand once.)
- Readable results. Text is now split into shorter chunks of about 90 words that end on sentence boundaries. Each result card shows its whole chunk, and clicking it highlights exactly that text in the reader, centered on screen. Near-empty fragments such as lone page numbers are no longer indexed.
- Faster upgrade. Existing libraries re-chunk only their text in the background on first launch; page images, audio and video are kept.
- Hyphenated words in queries.
-subtracts a term only when it starts a word, so queries likeleft-wingorCOVID-19work as expected. - PDF reader: a result near the bottom of a page now opens on the right page instead of the next one.
- Media timeline: lanes are now labeled Audio and Visual.
What's new in 0.2
- Multimodal search: one embedding space covers text, images, rendered PDF pages, audio (15 s windows) and video (frames plus soundtrack).
- Multimodal queries: search with text, an image or audio file, or your voice.
- New viewers: an image viewer with zoom and pan, and an audio/video player with a waveform, hit markers and seeking. Results show media thumbnails.
- Show filter: limit results to Text, Pages, Images, Video or Audio.
- Reliability: a bad PDF page no longer fails the whole PDF. Startup errors show an alert and are logged to
~/Library/Logs/Semantra/startup-error.log.
Full Changelog: https://github.com/freedmand/semantra/compare/v0.1.3...v0.2.2
