Understanding how climate change impacts berry ripening physiology is essential for selecting genotypes that balance sugars and acids under warming conditions. In this context, we used a portable near-infrared spectrometer in the vineyard, to monitor sugar and acid evolution in individual berries from 10 grapevine varieties over two years. Spectra were periodically acquired on the same berries all along ripening, and a subset of these berries was also collected for sugars and organic acids quantification by HPLC, to train partial least square regression models. Prediction models for glucose, fructose, and malic acid concentrations were fairly accurate with validation R2 of 0.71, 0.64, and 0.55, respectively. We further used these models to study sugar accumulation in individual berries and observed that the single berries ripen two times faster than average samples. Our results pave avenues toward precise quantitative approaches on sugar and acid fluxes in berry ripening studies.