The implementation of large models in the financial sector faces two major issues: one is the question of dominance, the competition between tech companies and financial institutions. The other is the challenge of integration, how to merge large models with financial operations. Financial institutions are concerned about missing the wave of large models but also face challenges such as data privacy compliance and other hard metrics. The article suggests that financial large models should prioritize suitable scenarios, and scenarios that are easy to deploy hold potential. However, the technology of large models itself still has issues, such as the black box nature of models and hallucination problems, which need to be overcome. The article believes that bridging the gap between industry and large models will be a future challenge, and early explorers of large models in financial institutions may provide valuable practical experiences.