Make scanned PDFs searchable and editable. Specially optimized for English + Marathi. Existing PDF text is detected first for speed; only scanned pages are sent through OCR — 100% free, 100% private. स्कैन की गई PDF को सर्च करने योग्य और एडिट करने योग्य बनाएं। इंग्लिश + मराठी के लिए खास तौर पर ऑप्टिमाइज़्ड। मौजूदा PDF टेक्स्ट पहले पहचाना जाता है; सिर्फ स्कैन किए पन्ने OCR से गुज़रते हैं — पूरी तरह मुफ्त, पूरी तरह प्राइवेट।
Tap to upload or drag your PDF/image hereअपलोड करने के लिए टैप करें या PDF/इमेज यहाँ खींचें
PDF, JPG, PNG
Use fewer languages for faster and more accurate recognition.तेज़ और सटीक पहचान के लिए कम भाषाएं चुनें।
Fast: selectable pages are skipped; scanned pages use lower render quality.फास्ट: पहले से सिलेक्टेबल पन्ने छोड़ दिए जाते हैं; स्कैन पन्ने कम क्वालिटी में प्रोसेस होते हैं।
Search is exact by default. Even one Marathi character or one English letter can be searched.सर्च डिफ़ॉल्ट रूप से एग्जैक्ट है। एक मराठी अक्षर या एक इंग्लिश अक्षर भी खोजा जा सकता है।
You can directly correct names, Marathi letters, and numbers here. Editing a scanned PDF's visible image itself requires a full PDF image editor; this editor changes the OCR/search text index.आप यहाँ सीधे नाम, मराठी अक्षर और नंबर ठीक कर सकते हैं। स्कैन की गई PDF की दिखने वाली इमेज एडिट करने के लिए पूरा PDF इमेज एडिटर चाहिए; यह एडिटर सिर्फ OCR/सर्च टेक्स्ट इंडेक्स बदलता है।
Scanned PDFs — old government forms, enumeration sheets, printed records — often can't be searched or copied because they're really just page images. This tool runs OCR (optical character recognition) directly in your browser, detecting pages that already have selectable text and skipping them for speed, then recognizing text on scanned pages using Tesseract.js with support for English, Hindi, Marathi, and Urdu. The result is a searchable PDF with an invisible text layer laid exactly over the original page image, so the document still looks the same but every word becomes findable and copyable. Nothing is uploaded to a server; the entire process, from reading the PDF to building the searchable output, happens on your device.
Once OCR finishes, every page's recognized text is indexed for instant search. The search box performs an exact match first — useful for finding a specific EPIC number, a single Marathi letter, or a name spelled a particular way. If nothing matches exactly, the tool checks a small set of characters that OCR commonly confuses, such as O and 0, I and 1, or S and 5, and shows those as clearly labeled possible matches rather than exact results. This two-step approach avoids false confidence: you always know whether a result is a confirmed exact match or a suggestion worth double-checking against the original scan.
Fast mode skips pages that already have selectable text and renders scanned pages at a lower quality for quicker OCR — a good default for long documents. Balanced and High Accuracy render scanned pages at higher resolution before recognition, which improves accuracy on small or low-contrast text at the cost of more processing time. After OCR completes, the built-in editor lets you correct misread names, numbers, or Marathi characters page by page — these corrections update the searchable text index immediately, so search results reflect your fixes even though the visible scanned image itself stays unchanged.