Bring in your files
Import documents, PDFs, images, and audio without sending the originals to a service.
OmniPrism Studio
Find the evidence you already own.
OmniPrism Studio is a local-first desktop library for searching across documents, scans, and recordings while keeping your processing and source data on your device.
Local processing
Original evidence retained
No remote parser
Search your libraryCtrl K
All evidenceDocumentsRecordings
DOC
Project handover notesOriginal source · 12 matches
0.94
WAV
Interview recordingTimestamped transcript · 8 matches
0.88
OCR
Signed invoice scanPage 2 · visual evidence retained
0.81
A different kind of search
Retrieval that points back to what was actually there.
OmniPrism Studio helps you find relevant passages, pages, and moments. It keeps the original source, timestamps, and review trail attached so the result remains something you can inspect—not an answer you have to take on faith.
How it works
From messy files to dependable evidence.
One desktop workflow for the material that normally gets scattered across folders, apps, and recordings.
Process locally
OCR, transcription, and semantic indexing run on your device using the processing profile you choose.
Search naturally
Combine exact terms with meaning-based retrieval, with optional refinement when accuracy matters most.
Review the source
Open the original context, verify the passage, and keep accepted evidence immutable and traceable.
Built for real collections
Useful across the formats you already have.
Start with the files on your computer. Add structure only where it helps you find and verify the underlying material.
TXT
Documents & PDFs
Plain text, Markdown, HTML, and PDF content stay connected to the source document.
OCR
Scans & images
Extract text from visual documents while retaining page context and normalized evidence anchors.
AUDIO
Recordings
Import WAV, MP3, and FLAC files with timestamped transcription evidence and original audio preserved.
EXP
Portable library
Export your accepted library and restore it later without depending on a remote account or service.
Language support
Interface in your language. Processing beyond it.
The OmniPrism Studio interface is available in 18 languages. Your documents do not need to match the language you choose for the app.
Available UI languages
Choose any of these languages in Settings. Names are shown in their native form.
EnglishالعربيةDeutschEspañolFrançaisहिन्दीBahasa Indonesia日本語한국어NederlandsPolskiРусскийไทยTürkçeTiếng ViệtPortuguês (Brasil)简体中文繁體中文
Broader processing language support
OCR, audio transcription, and semantic search rely on installed processing models. Those models support many languages beyond the interface, so you can work with multilingual collections while keeping the UI in the language you prefer.
01 OCR
02 Audio transcription
03 Semantic search
Coverage varies by model, file type, and installed resources.
Your data stays yours
Privacy is an architecture decision, not a settings page.
OmniPrism Studio is designed around local processing, user-controlled storage, and evidence you can inspect. Your library is not uploaded for search, while Store activity, model downloads, and support messages follow their own network paths.
Read YYamin's principlesOn-device by default
Processing is performed locally on the desktop.
Export when you choose
Your accepted library is not trapped in the app.
Evidence over invention
Search surfaces source material instead of generating a replacement answer.
Availability
Windows first. macOS follows after validation.
The Windows release is being prepared for packaging and store delivery. macOS implementation is in progress, with final testing intentionally scheduled after the rest of the release path is in place.
Ask about release updates