This paper presents TotalBotWar, a new pseudo real-time multi-action\nchallenge for game AI, as well as some initial experiments that benchmark the\nframework with different agents. The game is based on the real-time battles of\nthe popular TotalWar games series where players manage an army to defeat the\nopponent's one. In the proposed game, a turn consists of a set of orders to\ncontrol the units. The number and specific orders that can be performed in a\nturn vary during the progression of the game. One interesting feature of the\ngame is that if a particular unit does not receive an order in a turn, it will\ncontinue performing the action specified in a previous turn. The turn-wise\nbranching factor becomes overwhelming for traditional algorithms and the\npartial observability of the game state makes the proposed game an interesting\nplatform to test modern AI algorithms.\n