Opinion: Decentralized AI has changed because of Deepseek, and it is expected that AI models developed based on the complete Web3 AI stack will continue to emerge

Odaily News Pandu Fund released a research report titled "Decentralized AI is Changing Because of Deepseek", which pointed out that the decentralized AI narrative is constantly being reshaped. Web3 AI companies can focus on replicating DeepSeek's success while providing new advantages such as multimodality, user ownership, anti-censorship and privacy. It is expected that the number of projects on the supply side will continue to grow, and consumer-oriented projects will also begin to compete with Web2 competitors by building community-engaged networks. In the coming year, AI models developed based on the complete Web3 AI stack will continue to emerge.
In addition, companies that combine AI and encryption are gradually adjusting their strategies to focus on infrastructure construction rather than model development. For example, companies in the GPU market such as Akash, Render, IoNet, and Exabits have developed sustainable revenue models. At the same time, companies like Grass and Gradient, which allow users to share network bandwidth, have found a market position in providing services such as distributed networks to Web2 customers. In inference tasks, the performance gap between small and large models is narrowing, which means that Web3 does not need to rely on the super-large computing power of traditional AI giants to use these streamlined models for efficient inference operations. As this trend develops, we may see more inference endpoints driven by decentralized computing networks in the future.