Based on unique data from a worldwide survey among participants of international climate conferences, I investigate the acceptance of the most discussed components of architectures for an international climate agreement, namely: global quantitative targets, sector targets, research and development, geoengineering, land use, and adaptation. Regional and economic differences as well as personal attitudes play an important role for the perception of the different components. Global quantitative targets and adaptation are the most accepted in contrast to a low acceptance of geoengineering. People that are more affected by climate change and value fairness a lot care more about global and sector targets and research and development. Surprisingly, being vulnerable to climate change does not increase the preference for adaptation by much. Furthermore, I analyze which countries or groups of countries are expected to play a leading role for each component. The EU is seen as a key player and not much is expected from the USA and China. I detect a normative bias that increases expectations on China, the EU, and the USA for some of the components.