GLM-OCR
Optical character recognition powered by THUDM/glm-ocr-2b-1.1 via ZeroGPU. Upload an image, get the text out.
Upload a screenshot, photo of a document, whiteboard, or handwritten note. The model extracts all text it can find.
Examples
Upload a form, table, receipt, invoice, or business card. Pick an extraction type and get structured JSON output.
Extraction type
Examples
Upload up to 5 images. Each is processed separately and results are combined below.
Built by Xavier. Support the infra at ko-fi.com/xavierfuentes