Central to human sociality, evaluation occurs at personal, interpersonal, group, and intergroup levels, with competing theories and evidence. Five current social-evaluation models engage here in adversarial alignment, to identify common conceptual ground, ongoing controversies, and continuing agendas for work on social evaluation: Dual Perspective Model (Abele & Wojciszke, 2007); Behavioral Regulation Model (Leach, Ellemers, & Barreto, 2007); Dimensional Compensation Model (Yzerbyt, Provost & Corneille, 2005); Stereotype Content Model (Fiske, Cuddy, Glick, & Xu, 2002); and Agency-Beliefs-Communion Model (Koch, Imhoff, Dotsch, Unkelbach, & Alves, 2016). Each model has distinctive focal domains, theoretical roots, premises, and evidence. Controversies arise on the number, definition, and labeling of dimensions, their organization, their relative priority, and their relationship. A new alignment addresses these controversies. Two fundamental dimensions are Vertical (agency, competence, “getting ahead”) and Horizontal (communion, warmth, “getting along”); they respectively have facets of ability and assertiveness (Vertical) and friendliness and morality (Horizontal). Depending on context, a third dimension is conservative versus progressive Beliefs. Dimensions’ priority depends on the criterion (processing speed, pragmatic diagnosticity, subjective weight) and on moderators. Relationships between dimensions are predictably positive, orthogonal, negative, or curvilinear, depending on specific comparisons and task demands. Functions vary (e.g., comparison, accuracy, esteem, interaction) and depend on identifiable contexts. Processes (catalysts, antecedents, and consequences) depend on degree and types of interdependence. Integrating the models, this adversarial alignment generates new research directions and pragmatic functional implications.