China's DeepSeek 4.0 AI model finally released: How it compares with ChatGPT and Claude
Last Updated: 2026-04-24 16:36:03
DeepSeek, the Chinese artificial intelligence startup, has launched preview versions of its latest major update, V4. This release comes as the AI rivalry between China and the US continues to escalate. The company's previous offerings had already made an impact on global markets.
The DeepSeek V4 model has been keenly anticipated by users eager to assess its capabilities against leading US competitors such as OpenAI's ChatGPT, Anthropic's Claude, and Google's Gemini. The launch also follows earlier expectations that the new model might have arrived sooner, around the Lunar New Year.
Latest and Popular Mobiles
Enhanced capabilities and performance claims
DeepSeek states that its new V4 open-source models, including "pro" and "flash" versions, feature significant improvements across several key areas. These enhancements include advancements in knowledge acquisition, reasoning abilities, and their "agentic" capabilities, which refer to the AI's capacity to perform complex, autonomous tasks.
Related Articles
V4 serves as the successor to DeepSeek's V3, an AI model introduced in late 2024. However, it was the specialized "reasoning" AI model, R1, released in January 2025, that garnered significant attention. DeepSeek had claimed R1 was more cost-effective than similar models from OpenAI, positioning it as a symbol of China's technological progress.
The company asserts that its "V4 Pro Max" version delivers "superior performance" in standard reasoning benchmarks when compared to OpenAI's GPT-5.2 model and Google's Gemini 3.0-Pro. DeepSeek also notes that it falls only "marginally" short of GPT-5.4 and Gemini 3.1-Pro.
Regarding "agentic" capabilities, DeepSeek claims its V4 "pro" version outperforms Claude's Sonnet 4.5 and approaches the level of Claude's Opus 4.5 model, based on its internal evaluations. The "flash" version of V4 reportedly performs similarly to the "pro" version on simple agent tasks and has closely comparable reasoning capabilities.
Industry reactions and open source approach
Lian Jye Su, chief analyst at Omdia, commented that "it does appear DeepSeek V4 is going to be very competitive against its US rivals." Marina Zhang, an associate professor at the University of Technology Sydney, described the V4 rollout as a "pivotal milestone for China's AI industry," particularly given the intensifying global pursuit of technological self-reliance.
DeepSeek provides a free-to-use web and mobile chatbot. Unlike some top models from Anthropic, Google, and OpenAI, DeepSeek designates its technology as "open source," allowing developers to access, modify, and build upon its core framework. Both V4's "pro" and "flash" versions boast a 1 million token context window, a significant improvement from V3's 128,000 token window.
Skepticism and allegations
Despite the advancements, some analysts express skepticism. Ivan Su, a senior equity analyst at Morningstar, views V4 as a "competent" follow-up but not as significant a breakthrough as the R1 release. He also highlights intensified domestic competition since R1's introduction and emphasizes the need for independent evaluations to draw definitive conclusions about V4's capabilities against US models.
The release of V4 occurs amidst ongoing allegations of intellectual property issues. In February, Anthropic accused DeepSeek and two other China-based AI laboratories of "industrial-scale campaigns" to illicitly extract capabilities from its Claude models. OpenAI has made similar allegations in communications with US lawmakers, citing distillation techniques.
Michael Kratsios, former chief science and technology adviser to US President Donald Trump, also accused foreign tech companies, primarily based in China, of exploiting American AI expertise. China's embassy in Washington has refuted these allegations, characterizing them as "unjustified suppression of Chinese companies by the US."