Image OCR without a browser tab
Drop in a screenshot, clipboard image, or file and extract text from the desktop app.
Local-first desktop OCR
Convert images to text on your desktop.
pixtotext extracts text from screenshots, scanned snippets, and image files. OCR runs locally by default, with batch extraction, document conversion, offline translation, and history built in.
Available for Windows, macOS, and Linux through itch.io. Name your own price.
OCR runs locally by default No cloud upload unless you enable a cloud provider yourself.
Private first Local engines are the default path for image-to-text work.
Batch ready Process multiple images and merge results when the job calls for it.
Conversion built in Move OCR output into HTML, DOCX, PDF, or Markdown.
Optional cloud OCR Use higher-accuracy providers only when you intentionally configure them.
A focused tool for everyday OCR work.
pixtotext is built like a desktop utility: practical controls, real local engines, and enough depth for repeated OCR tasks.
Batch extraction
Queue multiple files, merge output when useful, and keep repetitive OCR work moving.
Convert documents
Turn OCR output into HTML, Word, PDF, or Markdown without leaving the app.
Offline translation
Translate extracted text locally with Argos Translate when language packs are installed.
History that stays useful
Review previous OCR work, reopen results, and clear history when you want a fresh slate.
Pick the OCR engine that fits the file.
Start local, stay local, and opt into cloud processing only for cases where you need it.
| Engine | Privacy | Best for | Notes |
|---|---|---|---|
| Native OCR | Local | Fast desktop OCR on supported operating systems. | Uses the platform OCR engine when available. |
| PaddleOCR | Local | Complex layouts and bundled offline recognition. | Useful when native OCR is unavailable or layout accuracy matters. |
| Tesseract | Local | Standard OCR jobs with a widely used engine. | Available when Tesseract is installed. |
| ocrs | Local | Small English text and early-preview local OCR. | Fast, lightweight, and still marked as a preview engine. |
| Cloud OCR | Optional cloud | Higher accuracy cases that justify sending data to a provider. | Only runs after you turn it on and add your provider details. |
Workflow
From image to usable text in three moves.
The app keeps single screenshots quick, while still handling larger batches when the job grows.
Bring in an image
Paste from the clipboard, pick a file, or use the desktop workflow that fits your OS.
Choose the OCR path
Run local engines by default, or opt into cloud OCR for tougher documents.
Use the text
Copy results, batch files, convert documents, translate offline, or return to history later.
Real product screens, not mockups.
See the actual desktop app: extraction, settings, batch work, conversion, and history.
FAQ
Straight answers before installing.
Does pixtotext upload my images?
No. OCR runs locally by default. Images are only sent to a cloud provider if you enable and configure cloud OCR.
What platforms does it target?
pixtotext is available for Windows, macOS, and Linux. Mobile builds are planned later.
Can it process more than one image?
Yes. The batch extractor is built for multiple files and can merge output when that helps the task.
Where does cloud OCR fit?
Cloud OCR is optional. It is meant for higher-accuracy cases where you intentionally choose a provider and accept sending data to that provider.
Install pixtotext and keep OCR close to your files.
Get the desktop build for your platform and start with local OCR by default.