DeepSeek, a generative AI (genAI) chatbot that originated from China, astonished the world with its capability to match ChatGPT while costing significantly less to build and using less powerful GPUs. The launch has triggered an industry soul search that could potentially accelerate AI adoption worldwide.

The cost efficiency underscored by DeepSeek has disrupted AI titans and emboldened smaller players

The DeepSeek chatbot, powered by its flagship R1 and V3 models, has shown the way for less resource-intensive large language models (LLMs). Using the model distillation technique—where smaller models are trained to mimic the performance of larger ones—developers have cut costs without sacrificing much performance.

In contrast to OpenAI, DeepSeek has adopted an open-source model—making the code free for anyone to inspect, download, and modify. DeepSeek has already prompted OpenAI to offer more of its services without charge.

The newfound cost efficiency could enable AI to thrive further in sectors and applications. DeepSeek has already inspired players with smaller budgets to establish LLMs, which may lead to the democratization of AI.