VIP pricing

← All tools

OCR text recognition

Chinese printed-text recognition, outputting plain text, JSON (with coordinates and confidence), or XML. Local recognition by default; switch to cloud for speed with many documents.

Drag in images or PDF, or click to choose
PNG / JPEG / WebP / PDF · multi-select
Recognition mode

How to use

  1. Drop in images or PDFs (batch supported)
  2. Pick the language and DPI; local recognition is the default (no upload)
  3. Copy the text, or export plain text / JSON with coordinates and confidence / XML

FAQ

Why download a model first?
Local recognition needs a ~44 MB Chinese model, downloaded once and cached — afterwards it even works offline.

How do I improve accuracy?
Use 300 DPI, keep text level, crop irrelevant areas. For many documents, switch to cloud recognition for speed.

Everything runs locally in your browser — files are never uploaded. BoxKit VIP unlocks unlimited use of all 29 tools.