In recent years, Smart Contracts have gained in popularity, facilitating billions of US Dollars in daily transactions. However, the recent increase in smart contract vulnerabilities threatens to undermine trust in the technology. The study aims to detect and address potential vulnerabilities in smart contracts in blockchain technology through a comprehensive analysis of four principal modalities: Solidity source code, bytecode, opcode, and intermediate representations. This proactive identification of vulnerabilities can contribute to bolstering the security and dependability of blockchain-based systems. In this paper, we propose a novel multimodal Transformer architecture named QuadraCode AI, utilizing these four distinct modalities. Unlike traditional unimodal analysis, multimodal analysis can provide a more holistic understanding of both the semantic and syntactical contexts of smart contracts to identify underlying vulnerabilities. By employing advanced data fusion techniques such as cross-attention and concatenations across 12 different multimodal frameworks, our approach enhances the detection capabilities beyond traditional unimodal approach. Notably, the framework that integrates opcode with bytecode achieves an impressive average F score of 86%, demonstrating the effectiveness of our method.