HONG KONG (AP) 鈥� DeepSeek, the Chinese artificial intelligence startup that shook world markets last year, launched preview versions of its latest major update Friday as the AI rivalry between China and the U.S. heats up.
DeepSeek鈥檚 V4 has been keenly anticipated by users looking to test how it compares to U.S. competitors like OpenAI鈥檚 ChatGPT, Anthropic鈥檚 Claude and Google鈥檚 Gemini. Anthropic and OpenAI have accused DeepSeek of unfairly building its technology off their own.
Some industry analysts had expected the new model to arrive more than two months earlier at the start of .
DeepSeek says the new V4 open-source models, which include 鈥減ro鈥� and 鈥渇lash鈥� versions, have big improvements in knowledge, reasoning and in their 鈥渁gentic鈥� capabilities 鈥� the ability to perform complex tasks and workflows autonomously. Another big change is they are supported in part by computer chips made by Chinese tech giant Huawei, reducing DeepSeek’s reliance on U.S. chipmakers like Nvidia.
V4 is a successor to V3, an AI model that DeepSeek released in late 2024.
But it was DeepSeek鈥檚 specialized 鈥渞easoning鈥� AI model, , that took markets by surprise with its release in January 2025. DeepSeek claimed it was more cost-effective than OpenAI鈥檚 similar model and it became a symbol of how China was with the U.S. in technological advancements.
DeepSeek said the 鈥淰4 Pro Max鈥� version has 鈥渟uperior performance鈥� in terms of standard reasoning benchmarks relative to OpenAI鈥檚 GPT-5.2 model and Google鈥檚 Gemini 3.0-Pro. It falls 鈥渕arginally鈥� short of GPT-5.4 and Gemini 3.1-Pro, it said. DeepSeek’s release came hours after its new GPT-5.5 model on Thursday.
In terms of 鈥渁gentic鈥� capabilities, the Chinese company said the V4 鈥減ro鈥� version could outperform Claude鈥檚 Sonnet 4.5 and approaches the level of Claude’s Opus 4.5 model based on its own evaluation.
The 鈥渇lash鈥� version of V4 performs on a par with the 鈥減ro鈥� version on simple agent tasks and has reasoning capabilities closely approaching it, DeepSeek said.
鈥淏ased on the benchmark results, it does appear DeepSeek V4 is going to be very competitive against its U.S. rivals,鈥� said Lian Jye Su, chief analyst at the technology research and advisory group Omdia.
Marina Zhang, an associate professor at the University of Technology Sydney, said DeepSeek’s V4 rollout is as a 鈥減ivotal milestone for China鈥檚 AI industry,鈥� especially as global competition intensifies in the pursuit of self-reliance in critical technologies.
DeepSeek offers a free鈥憈o鈥憉se web and mobile chatbot. Unlike the top models from Anthropic, Google and OpenAI, it describes its technology as 鈥渙pen source鈥� in the way that it enables developers access to modify and build on its core technology.
Both the V4’s 鈥減ro鈥� and 鈥渇lash鈥� versions have a 1 million token context window, a parameter of how much information an AI model can process and recall, and run on a more efficient basis, the startup said. That is a significant improvement from before, since the V3 supported a 128,000 token context window.
Huawei said in a separate statement Friday that its Ascend chips and related technology are compatible with the DeepSeek V4 models. It鈥檚 a demonstration of technical feasibility of operating outside the Nvidia-dominated computing ecosystem 鈥渁mid sustained technological decoupling between China and the U.S.,鈥� said Zhang.
A from Microsoft in January showed use of DeepSeek has been gaining ground in many developing nations, particularly those where Huawei phones are widely used.
However, some analysts remain skeptical. Ivan Su, a senior equity analyst at Morningstar, said while V4 is a 鈥渃ompetent鈥� follow-up, it鈥檚 not as big a breakthrough as the rollout of R1.
鈥淒omestic competition has intensified significantly since R1鈥檚 release,鈥� Su said. 鈥淎gainst U.S. models, DeepSeek鈥檚 own evaluation suggests its capabilities largely match on most fronts, but independent evaluations are needed before final conclusions can be drawn.鈥�
In February, Anthropic accused DeepSeek and two other China-based AI laboratories of 鈥渋ndustrial-scale campaigns鈥� to 鈥渋llicitly extract Claude鈥檚 capabilities to improve their own models.鈥� It said they did that using a technique called distillation that 鈥渋nvolves training a less capable model on the outputs of a stronger one.鈥� OpenAI made similar allegations in a letter to U.S. lawmakers.
This week, Michael Kratsios, chief science and technology adviser to U.S. President Donald Trump, also of distilling leading U.S. AI systems and 鈥渆xploiting American expertise and innovation.鈥�
China鈥檚 embassy in Washington hit back at the allegations, describing them as 鈥渦njustified suppression of Chinese companies by the U.S.鈥�
___
O鈥橞rien reported from Providence, Rhode Island.
Copyright © 2026 The Associated Press. All rights reserved. This material may not be published, broadcast, written or redistributed.
