100tools
AssetsBeta

OCR Text Recognition

Extract text from images locally in the browser with Paddle.js.

OCR Text Recognition

Protect privacy by recognizing text locally in your browser.

Drop an image here or choose a local file. Nothing is uploaded.

Selections0
ModelAI model loads before the first recognition.

No regions selected. The full image will be recognized.

Upload an image to start selecting recognition regions.

Recognition result

No text recognized yet.

Detected lines

Each selected region keeps its own recognition block.

No text recognized yet.

Related Tools

Image CropperCrop a precise region from an uploaded image and export a new asset.AssetsImage ConverterConvert uploaded images between common web and desktop formats.AssetsText ProcessorApply common text cleanup transforms like single-line conversion and whitespace normalization.Text & Data

Frequently Asked Questions

Does the image leave my browser?

No. Recognition runs in the browser with Paddle.js. Uploaded images and selected regions are processed locally and are not sent to the 100tools server.

What languages can it recognize?

The bundled Paddle OCR model is designed for Chinese, English, and numbers. Image quality, contrast, and text size still affect accuracy.

How do selected regions work?

Drag across the preview to create one or more rectangular regions. Click an existing region to remove it. If no region is selected, the full image is recognized.