Chinese startup DeepSeek has released a preview version of its new artificial intelligence model, V4, adapted to run on Huawei chips, according to multiple reports. This move marks another step in China's push to build a self-sufficient AI ecosystem.
Key Takeaways
DeepSeek has released preview versions of its V4 artificial intelligence model, adapted for Huawei chips, marking a step in China's push for an independent AI ecosystem. The model comes in two versions: Pro and Flash, both supporting a 1-million-token context window.
- DeepSeek-V4 is designed to work with agent frameworks like Claude Code and OpenClaw.
- V4-Pro outperforms all open-source models but trails Google's Gemini and OpenAI's GPT in some areas.
- V4-Flash offers faster response times at a lower cost than Pro.
- The model was adapted for Huawei's Ascend AI chips, reflecting China's effort to reduce reliance on U.S. technology.
- DeepSeek faces compute constraints under U.S. export controls, limiting current Pro service availability.
The model comes in two versions: the more powerful and expensive Pro, and the cheaper, lighter Flash. DeepSeek-V4-Pro is positioned as a higher-end model with performance comparable to leading closed-source systems, particularly in agentic coding, world knowledge, STEM (science, technology, engineering and mathematics), and competitive programming. In maximum reasoning mode, Pro outperforms all open-source models but still trails frontier closed-source systems such as Google's Gemini 3.1 Pro and OpenAI's GPT-5.4 in some areas.
DeepSeek-V4-Flash delivers similar reasoning ability in some areas but runs faster and at a lower cost than Pro, with weaker world knowledge and lower performance on more demanding agent-based tasks. Both versions support a 1-million-token context window, matching the expansion DeepSeek introduced with V3 in February.
A key change from earlier DeepSeek releases is that V4 was adapted for Huawei's most advanced Ascend AI chips. According to Reuters, hours after the preview release, Huawei said V4 is fully supported on its Ascend 950-based supernode clusters, and that its chips were used for part of V4-Flash's training.
The partnership between DeepSeek and Huawei shows progress toward AI infrastructure self-sufficiency. However, DeepSeek faces compute constraints under U.S. export controls on Nvidia chips and chipmaking equipment. The company said Pro can cost up to 12 times more than Flash because of 'constraints in high-end compute capacity,' limiting current Pro service availability.
How this summary was created
This summary synthesizes reporting from 3 independent publishers using AI. All sources are cited and linked below. NewsBalance is a news aggregator and media literacy tool, not a news publisher. AI-generated content may contain errors or inaccuracies — always verify important information with the original sources.