China’s DeepSeek previews new AI model a year after jolting US rivals

2026-04-24 · Source: The Verge · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Fundamental Awareness, quick

Summary

Chinese AI company DeepSeek has released a preview of its next-generation AI model, V4, claiming it can compete with leading closed-source systems from US rivals like Anthropic, Google, and OpenAI. DeepSeek states that V4 significantly improves over previous models, particularly in coding capabilities, which are crucial for AI agents and tools such as ChatGPT Codex and Claude Code. This release also marks a notable achievement for China's chip industry, as DeepSeek explicitly highlights V4's compatibility with domestic Huawei technology. The company previously gained attention with its R1 model, which it claimed was trained at a fraction of the cost of US systems, though V4's training costs and hardware remain undisclosed. DeepSeek has faced accusations of using banned Nvidia chips and misusing Anthropic's Claude to enhance its products.

Key takeaway

For AI Architects evaluating new foundation models, DeepSeek's V4 preview suggests a competitive open-source option, particularly for coding applications and those prioritizing compatibility with Huawei hardware. You should assess V4's performance benchmarks against established models like Anthropic's Claude and OpenAI's systems, especially if seeking alternatives to US-centric AI infrastructure. Be mindful of ongoing geopolitical and intellectual property concerns surrounding DeepSeek's training practices.

Key insights

DeepSeek's V4 model aims to challenge top US AI systems, emphasizing coding prowess and domestic chip compatibility.

Principles

• Open-source models can rival closed-source systems.
• Coding capability is central to advanced AI agents.

In practice

• Evaluate V4 for coding-intensive AI agent development.
• Consider V4 for projects requiring Huawei chip compatibility.

Topics

• DeepSeek V4
• Open-source AI
• AI Coding Capabilities
• China's AI Industry
• US-China Tech Rivalry

Best for: CTO, VP of Engineering/Data, AI Architect, AI Scientist, Director of AI/ML, Tech Journalist

Related on AIssential

• China’s DeepSeek Just Dropped a New AI Model Built on Huawei Chips — DeepSeek's V4 model demonstrates high-performance AI can be developed using domestic Chinese hardware, challenging U.S. chip dominance.
• AI News: Everyone's Mad At Anthropic Now — Open-source AI models are rapidly closing the performance gap with proprietary models while offering significant cost and privacy advantages.
• What’s next for Chinese open-source AI — Chinese open-source AI models offer near-frontier capabilities at a fraction of the cost, rapidly reshaping global AI innovation.
• Deepseek is a Problem — The US open-source AI model is failing due to lack of monetization, risking Chinese dominance and geopolitical influence.
• ChinAI #359: DeepSeek's "Huawei-like" Mission in AI — DeepSeek's soaring valuation reflects its strategic role in China's drive for AI self-sufficiency amidst external tech restrictions.
• My Honest Thoughts about Deepseek — DeepSeek V4 offers frontier-level AI capabilities at a fraction of the cost, challenging US AI dominance.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Verge.