The developers of GermanQuAD and GermanDPR have announced the release of two new datasets and three machine learning models in German, aiming to advance natural language processing (NLP) capabilities. The datasets, GermanQuAD and GermanDPR, are designed for question answering and dense passage retrieval tasks, respectively, and are based on human-annotated data. Two QA models, gelectra-base-germanquad and gelectra-large-germanquad, have been trained on these datasets and are available for download through Hugging Face's Model Hub. A third model, GermanDPR, has also been developed for dense passage retrieval tasks using the same dataset. The development of these resources is open-source, aiming to facilitate the creation of semantic search technologies in German.