In this article we propose an analysis of three Python learning environments : AlgoPython, Citizen Code Python and Pyrates. First, we present the functioning of each EIAH, the knowledge worked in these environments and the organization of this work. Then we study the hints and feedbacks proposed, and the orchestration possibilities. This work allows us to highlight the similarities and differences between these environments and raises various questions about their contributions in terms of learning. In conclusion, we synthesize these analyses and several possible uses of this work.