In a significant shift within the global AI market, DeepSeek, a Chinese artificial intelligence startup, has cut the price of its flagship V4 Pro model by 75%. This change highlights a transition in AI models from focusing solely on performance metrics to emphasizing cost efficiency, especially in competition with established U.S. tech giants.
According to the South China Morning Post, this price reduction, initially a promotional tactic, has now become the official price for the V4 Pro. This adjustment positions the model as a serious competitor against offerings from major players like OpenAI and Anthropic, delivering performance levels that can match them at a much lower cost.
Cost Efficiency Takes Center Stage
DeepSeek's pricing strategy emerges as the AI industry faces rising costs due to limited supplies of essential computing resources. The V4 Pro now has an API price as low as $0.0036 per million cached input tokens and $0.87 per million output tokens. Estimates from the benchmark firm Artificial Analysis indicate that running their “Intelligence Index” test with the V4 Pro costs about $268. This stands in stark contrast to OpenAI’s GPT-5.5, which incurs expenses approximately 12 times higher, and Anthropic’s Claude Opus 4.7, which is about 19 times more expensive.
The introduction of the V4 series, which includes the lighter V4 Flash model, underscores DeepSeek's commitment to aggressive pricing. The company has hinted at potential further cuts later this year, suggesting a continuing evolution in how AI models are priced and perceived.
A Shift in Competitive Standards
DeepSeek's pricing strategy has implications beyond its own products. Other Chinese firms, including MiniMax and Xiaomi, are also entering the cost-efficiency race. MiniMax’s M2.7 and Xiaomi’s MiMo V2.5 Pro have attracted attention for their competitive pricing and performance, while Alibaba has joined the competition with a 50% reduction in its latest model, Qwen3.7 Max.
As Chinese models become more cost-effective, adoption barriers for AI technologies decrease, making advanced AI accessible to a broader audience. DeepSeek’s V4 Flash has launched at the same price point as the V2 model released two years ago, reflecting a trend of downward adjustments in pricing structures.
Future Outlook
Industry analysts predict that this pricing shift could disrupt the global AI ecosystem. The emerging standard focuses not just on achieving the highest performance but on delivering the most value for the cost. As competition intensifies, particularly among Chinese firms, traditional pricing models may soon become outdated.
DeepSeek's aggressive pricing strategy is reshaping how companies approach their pricing and performance benchmarks. The anticipated release of Huawei Technologies’ Ascend 950PR supernode could further influence pricing strategies, potentially leading to additional reductions and shaping the future of AI pricing worldwide.
DeepSeek's bold reduction in V4 Pro pricing signals a key moment in AI, where cost efficiency may take precedence over performance, challenging established norms and building a more competitive market environment.