China's AI price war is exposing India's biggest weakness: Why India needs its own DeepSeek

China's artificial intelligence race is no longer just about building more powerful models—it is increasingly about making them dramatically cheaper. That shift could have significant implications for India, where experts warn the country risks becoming dependent on foreign AI models even as China rapidly strengthens its position in the global AI ecosystem.

China's low-cost AI push

A recent report by Jefferies describes the launch of GLM-5.2, developed by Hong Kong-listed Z.ai (formerly Zhipu AI), as another "DeepSeek moment." According to the brokerage, the model delivers enterprise-grade performance that is close to Anthropic's leading AI systems while costing only about one-quarter as much per token. The lower cost, Jefferies says, could become a decisive advantage as businesses increasingly choose AI models based on economics rather than marginal gains in performance.

That price advantage is already beginning to influence adoption. Jefferies cites data from AI platform OpenRouter showing that Chinese AI models processed 21.37 trillion tokens during the week ended June 21, compared with 5.76 trillion tokens for the leading US models. The brokerage argues that as large language models become increasingly commoditised, enterprises will prioritise affordability, deployment flexibility and data security over small differences in benchmark performance.

MUST READ: Forget ChatGPT? China's cheaper AI models are quietly taking over enterprise AI, says Jefferies

India's AI challenge is bigger than ChatGPT

For India, however, the challenge goes well beyond competing with OpenAI or Anthropic.

In a separate report, Bernstein argues that India has yet to experience its own "DeepSeek moment." While the country has built a globally competitive IT services industry and is witnessing rapid growth in AI applications, it still lacks a globally competitive foundational large language model. That dependence on foreign AI platforms, the brokerage warns, could become a strategic vulnerability if geopolitical tensions or export controls restrict access to advanced AI technologies.

MUST READ: India needs its own DeepSeek to avoid AI dependence, warns Bernstein

Bernstein points out that recent restrictions on access to some frontier AI models for non-US users underline a broader trend. Artificial intelligence is increasingly being treated as a strategic national asset alongside semiconductors, advanced chips and defence technologies. The brokerage compares frontier AI models to "fighter jets," arguing that countries are becoming less willing to freely share their most advanced capabilities.

The risk of relying on foreign AI models

This raises an important question for India. If enterprises, government agencies and even defence applications rely primarily on foreign AI models, what happens if access is delayed, restricted or becomes significantly more expensive?

Bernstein warns that India could end up operating one or two generations behind global competitors, reducing the competitiveness of its software industry despite having one of the world's largest engineering talent pools. It argues that India's technology ecosystem evolved around IT services and application development, while China built domestic internet giants that generated vast proprietary datasets, AI research capabilities and engineering talent. Those structural advantages are now helping Chinese companies build globally competitive AI models at lower costs.

MUST READ: India attracts AI billions, but HSBC's top investment picks are China, Japan and Korea; know why

A roadmap for India's AI future

That does not necessarily mean India must immediately replicate OpenAI or DeepSeek. Bernstein believes India's biggest opportunity lies in developing domain-specific AI models built on proprietary datasets across sectors such as healthcare, manufacturing, financial services and industrial automation. Building sovereign AI capabilities in these specialised areas could reduce dependence on foreign platforms while creating globally competitive products.

Taken together, the Bernstein and Jefferies reports suggest the next AI battle may not be decided solely by who builds the smartest model. China's strategy of combining improving model quality with aggressive pricing is already reshaping enterprise AI economics. For India, the bigger question is whether it continues to focus primarily on AI applications or begins investing more aggressively in sovereign foundation models. Without that shift, India risks remaining a consumer of AI technologies developed elsewhere while China increasingly becomes a global supplier of affordable enterprise AI.

MUST READ: ‘I handed Claude my salary’: Viral X post shares 7 prompts for financial freedom