This paper examines how parties frame their thematic and programmatic priorities in the run-up to the German state elections in Baden-Württemberg and Rhineland-Palatinate in early 2026. It makes a systematic comparison between the election programs of the major parties in the two federal states, the press releases of the state associations, and social media posts by the leading candidates. The theoretical background of this paper is formed by the perspectives of political discourse analysis on the one hand and framing perspectives on the other, so that language is viewed as an instrument for strategic use and functions as a means of constructing a desired meaning. Furthermore, language is recorded along the axes straightforward to complex and weak to powerful in order to reveal differences in argumentative density, linguistic strength, and communicative complexity and to enable conclusions to be drawn about the group of addressees. In addition, an ideal-typical framework is used that operationalizes conservative, liberal, social democratic, ecological, and populist language as programmatic orientation. Quantitative text methods such as natural language processing, topic modeling, semantic distance measures, and issue-specific lexicons are used to analyze the corpus. The lexicons were expanded and validated with the help of various large language models and finally checked for quality. Despite their considerable potential, the integration of LLM models is limited. Furthermore, the analysis of emotional expressions and narrative patterns marks the final necessary step in the development of a model that classifies the corpora based on their discursive or ideological communication style.