OCR
Image to text, local OCR.
Upload, drop or paste an image/screenshot to recognize its text. Powered by tesseract.js (WASM) running locally — images are never uploaded to a server. First use downloads a model from a CDN (Chinese ≈ 20MB), please wait.
Everything runs locally in your browser
Does it run locally?
Yes. tesseract.js compiles the engine to WebAssembly that runs in your browser, so images and results never leave your device. Only the engine and the chosen language model are fetched from a CDN on first use, then cached.
FAQ
How accurate is it?
Clear, high-contrast images with regular fonts work best; handwriting, blur, skew or decorative fonts reduce accuracy. Crop, enlarge and sharpen first.
Why is the first run slow?
It downloads the WASM engine and language model (Chinese is large) the first time; afterwards it’s cached and much faster.
Are my images safe?
Yes. All recognition is local; images are never uploaded or stored.
使用步骤How to use
- 上传或粘贴含文字的图片Upload or paste an image containing text
- 选择识别语言（中英混合等）Choose the recognition language
- 识别完成后复制文字Copy the recognized text