SoloScanDownload

Release notes

Changelog

Every meaningful change to SoloScan, newest first.

v0.1.6

Beta

New

  • Feedback from inside the app: a quick, optional way to tell us how SoloScan is working for you. It's private by default, and you only share a public testimonial if you choose to.

Improvements

  • Cleaner text extraction on scanned documents: fixed an issue where some scans picked up stray characters from the wrong language. Latin-script documents now read without that noise, and Arabic scans are still handled.
  • The website link in Settings now opens in your browser.

Under the hood

  • Scanned-document OCR now picks the language set per document instead of running every bundled language at once, which is what caused the stray characters.

v0.1.5

Beta

New

  • Vision model: when you analyze a scan or photo, SoloScan can now read the image directly instead of only its extracted text. This gives noticeably better answers on receipts, photos, handwriting, and multilingual documents that text reading struggles with.
  • Text/Vision switch: on scanned images, a simple toggle in the chat lets you choose how SoloScan reads the document, with a short note explaining the difference.
  • Free plan: SoloScan is now free to download and use forever. The free plan analyzes the first 1,000 characters of any document, so you can confirm it reads your real files before you pay. A license removes the limit and unlocks Vision.

Improvements

  • Answers are more careful: SoloScan now flags uncertainty instead of stating a shaky guess as fact, and stays focused on your document rather than drifting off-topic.
  • Responses no longer run on indefinitely on difficult scans; SoloScan stops cleanly when the answer is complete.
  • The chat now follows along to the newest text as a reply streams in, and stops following the moment you scroll up to re-read.

Fixes

  • Scanned documents with non-Latin numerals (such as Arabic-Indic digits) are handled more reliably and no longer drop real numeric lines.

Under the hood

  • The vision model now receives the actual image data, not just OCR text, so its image-reading capability is genuinely used.
  • The 7-day trial was replaced with a free plan whose limit is structural, which is simpler and fairer.

v0.1.4

Beta

New

  • Six-language document reading: English, Swedish, German, French, Spanish, and Arabic. Documents that mix scripts and numbers on the same line are now read in the correct order.
  • Copy button on the extracted-content panel, copy the full extracted text in one click.
  • Back button when you open a past scan from the Scans list, returning you to the same place in the list.

Improvements

  • Right-to-left text (Arabic) now displays in the correct reading order in both the extracted-content panel and the chat.
  • Extracted text keeps its line structure instead of collapsing into one block, so invoices and statements stay readable.
  • PDF text extraction now tries two engines and keeps the cleaner result, recovering text from documents that previously came out garbled.
  • Stop button during a response now stops immediately instead of waiting for the model to finish.
  • The chat scrolls to your new message when you send, then leaves you free to scroll while the reply streams in.

Fixes

  • OCR language data is now bundled with the app, so scanned documents read correctly on every machine, not just the developer's.
  • If Ollama isn't running when you send a message, your text stays in the box instead of disappearing.
  • The chat input no longer shows a scrollbar when it only holds a line or two.

Under the hood

  • Switched scanned-document OCR to bundled language data (English, Swedish, German, French, Spanish, Arabic) plus orientation detection, so OCR no longer depends on anything installed separately.
  • Replaced the single PDF text extractor with a two-engine approach that scores each result and keeps the cleaner one.

v0.1.3

Beta

New

  • Start Ollama for me, a one-click button that launches Ollama in the background when SoloScan needs it. No menu bar icon, no extra window, just ready to go.
  • Manage subscription directly from Settings → License. Opens the Stripe Customer Portal in your browser to cancel, update payment, or change seats.

Improvements

  • Reduced bottom whitespace across Home, Scans, and Settings so the page feels tighter and more deliberate.
  • Settings → About now shows the actual installed version at runtime instead of a stale build-time value.
  • More consistent layout when navigating between pages, no more subtle horizontal shift when entering Scans.

Fixes

  • Reopened scans now show the correct file size in their detail view, not 0 B.
  • File sizes in the Scans list are now real file sizes instead of being labelled as characters.

Under the hood

  • Auto-updates now flow through SoloScan's own infrastructure at updates.soloscan.app instead of a public hosting service. Same speed, better privacy, fully under our control.
  • Stored file size in the local database to support past-scan detail views going forward.

v0.1.2

Beta

Fixes

  • Fixed a packaging issue where the auto-updater would download and verify the new version but silently fail to install it. macOS metadata files in the update archive were invalidating the app signature post-extraction; the build now produces clean archives without that metadata.
  • Cleaned up update-related diagnostics that were occasionally shown to users.

v0.1.1

Beta

Improvements

  • Cleaner error messages throughout the app, no more technical jargon when something goes wrong.
  • Better typography on page titles using Space Grotesk for visual continuity with the wordmark.
  • Added click-to-browse alongside drag-and-drop on the file picker.
  • Bigger sidebar logo and more breathing room at the bottom of long pages.
  • Keyboard accessibility improvements across Settings and Scans. Tab navigation now reveals visible focus rings on every interactive element.

Fixes

  • Reopened scans now land at the top of the document, not the bottom of the chat.
  • Inner chat panel jumps to the most recent message on reopen, so you see where you left off.
  • UI chrome (logos, titles, navigation) is no longer text-selectable, matching native macOS expectations.
  • Re-bundled missing libraries that prevented the app from launching on Macs other than the developer's.

v0.1.0

Initial release

Highlights

  • First public build of SoloScan.
  • Local AI document analysis via Ollama (llama3.2 and llama3.2-vision).
  • Support for PDF, DOCX, plain text, and image files (PNG, JPG, WebP, TIFF, BMP).
  • OCR for scanned PDFs and image-only documents.
  • Multi-turn chat with persistence, past scans are saved locally and reopen instantly.
  • Light, Dark, and System theme support.
  • Signed and notarized by Apple for Gatekeeper-friendly install.