China's DeepSeek Unveils Shocking AI Update: 5 Game-Changing Features You Can’t Afford to Miss!

HONG KONG (AP) — DeepSeek, the Chinese artificial intelligence startup that caused waves in global markets last year, has launched preview versions of its highly anticipated V4 update. This release occurs amid intensifying rivalry between China and the U.S. in the AI sector, highlighting the increasing global stakes in technology development.

Users have been eagerly awaiting DeepSeek's V4, eager to see how it stacks up against American competitors such as OpenAI's ChatGPT, Anthropic's Claude, and Google's Gemini. Tensions are palpable, as both Anthropic and OpenAI have accused DeepSeek of unethically leveraging their technology to enhance its own offerings.

Initial expectations had pointed to a launch earlier in the year, coinciding with the Lunar New Year, but the new model has finally been unveiled. The V4 comes in several versions, including "pro" and "flash," boasting significant advancements in knowledge, reasoning, and "agentic" capabilities—this refers to the AI's ability to execute complex tasks and workflows autonomously.

DeepSeek's V4 is the successor to its V3 model, which was released in late 2024. The earlier model gained attention for its specialized “reasoning” AI, known as R1, which surprised markets when it was released in January 2025. DeepSeek claimed R1 was more cost-effective than similar models from OpenAI, signaling China's rapid technological progress compared to the U.S.

According to DeepSeek, the “V4 Pro Max” version demonstrates "superior performance" in standard reasoning benchmarks compared to OpenAI's GPT-5.2 and Google's Gemini 3.0-Pro, although it falls "marginally" short of GPT-5.4 and Gemini 3.1-Pro. In terms of agentic capabilities, DeepSeek asserts that the V4 “pro” version could outperform Claude's Sonnet 4.5 and is nearing the level of Claude's Opus 4.5 model based on its own evaluations.

The “flash” version of V4, while designed for quicker tasks, performs comparably in basic agent tasks and has reasoning capabilities that closely approach the "pro" version's abilities, according to the company.

Industry analysts are taking note; Lian Jye Su, chief analyst at Omdia, stated, “Based on the benchmark results, it does appear DeepSeek V4 is going to be very competitive against its U.S. rivals.” This sentiment is echoed by Marina Zhang, an associate professor at the University of Technology Sydney, who views the V4 rollout as a “pivotal milestone for China’s AI industry,” especially as global competition mounts in the race for self-reliance in critical technologies.

DeepSeek differentiates itself by offering a free-to-use web and mobile chatbot and promoting its technology as “open source,” which allows developers to modify and build upon its core technology. Notably, both the V4 "pro" and "flash" versions feature a 1 million token context window, significantly enhancing the amount of information the AI can process compared to the 128,000 token context window of V3.

A report from Microsoft in January indicated that DeepSeek's technologies are gaining traction in various developing nations, suggesting a growing influence in global markets. However, skepticism remains among some analysts. Ivan Su, a senior equity analyst at Morningstar, characterized V4 as a “competent” follow-up but not as groundbreaking as its predecessor, R1. He noted that domestic competition has increased substantially since R1's release, and while DeepSeek claims parity with U.S. models, independent evaluations are necessary for a definitive assessment.

In February, Anthropic accused DeepSeek and other Chinese AI labs of conducting “industrial-scale campaigns” to “illicitly extract Claude’s capabilities” to enhance their own models. This process, known as distillation, involves training a less capable model on the outputs of a stronger one. OpenAI has voiced similar concerns in a letter to U.S. lawmakers.

This week, Michael Kratsios, chief science and technology adviser to U.S. President Donald Trump, reiterated allegations that foreign tech companies, primarily those based in China, are distilling U.S. AI systems and exploiting American innovation. In response, China's embassy in Washington rejected the accusations, labeling them as “unjustified suppression of Chinese companies by the U.S.”

The unfolding developments in the AI sector reflect not just a technological arms race, but also a complex geopolitical landscape where both the U.S. and China are vying for dominance. As DeepSeek's innovations emerge, the implications for the global technology market could be far-reaching, raising questions about competition, ethical practices, and the future of AI development.