Exploring regional economic development has historically depended on data that oGen lags, is biased toward certain sectors, or is narrowly focused. Recent advancements in data collection and analysis, such as website scraping, provide fresher insights but also come with their own biases. Recognizing the importance and u:lity of national economic sentiment indices, this paper introduces a novel, location-specific, real-time indicator of economic sentiments derived from news data using the RegNeS database. This database collates daily news from over 300 German-language sources since 2019 that are geolocated. Economy related news articles are identified using deep learning methods. In addition to the thematic classification, sentiments are quantified by means of a multilingual BERT model, providing an unprecedentedly real-time view of regions’ economic development over time. The paper details the methodology of creating a regional economic sentiment index and evaluates its potential to capture the economic atmosphere of specific areas. Specifically, the index is compared to established economic sentiment indicators at the national level. It juxtaposes this index with traditional national economic sentiment metrics and examines its correlation with regional GDP growth through spatial econometric analysis. Overall, the research underscores the significance and u:lity of regional indices in deepening our comprehension of economic trends at a more localized level.