Data sharing and content offloading among vehicles is an imperative part of the Internet of Vehicles (IoV). A peer-to-peer connection among vehicles in a distributed manner is a highly promising solution for fast communication among vehicles. To ensure security and data tracking, existing studies use blockchain as a solution. The blockchain-enabled Internet of Vehicles (BIoV) requires high computation power for the miners to mine the blocks and let the chain grow. Over and above, the blockchain consensus is probabilistic and the block generated today can be eventually declared as a fork and can be pruned from the chain. This reduces the overall efficiency of the protocol because the correct work done initially is eventually not used if it becomes a fork. To address these challenges, in this paper, we propose a Directed Acyclic Graph enabled IoV (DAGIoV) framework. We make use of a tangle data structure where each node acts as a miner and eventually the network achieves consensus among the nodes. A game-theoretic approach is used to model the interactions between the vehicles providing and consuming offloading services. The proposed model is proven to be highly scalable and well suited for micro transactions or frequent data transfer among the nodes in the vehicular network.