Radar ·
Gemma 4 31B reads handwritten medical prescription in informal OCR test
A post from October 27, 2026 reports that the Gemma 4 31B model from Google read a handwritten medical prescription image better than earlier OCR models, in a test run by the author. It gives no link or benchmark, so the result is anecdotal.
A post published on October 27, 2026 reports that the Gemma 4 31B model from Google read a handwritten medical prescription image better than earlier OCR models, in a writing test run by the author. The post includes no link and no benchmark details, and the result rests on a single informal example.
The source itself says the finding is anecdotal and needs formal evaluation. To check it, find the original post by its date and title, look for a test set, metrics and methodology, and compare other OCR models on similar samples. Prescriptions contain personal and health data, so do not send real images for testing without authorization and prior anonymization.
Original source: Gemma 4 31B lê uma receita médica manuscrita em teste de OCR