According to Deutsche Bank Research, proprietary AI models may struggle to sustain their pricing premiums as low-cost open-source weight alternatives rapidly close the performance gap. The institution notes that as low-cost models continue to excel in benchmark tests, the economic viability of cutting-edge AI models is receiving increasing scrutiny.

Deutsche Bank believes the market is entering a phase akin to the one triggered by DeepSeek, where high-cost models face pricing pressure and enterprise clients reassess their substantial investments in proprietary models. This trend indicates that the competitive logic within the AI industry is shifting from pure performance pursuit toward cost-effectiveness, compelling leading vendors to adjust their pricing strategies in response to the growing impact of the open-source ecosystem.