China's DeepSeek says releases long-awaited new AI model
Summary
Chinese startup DeepSeek has launched a new AI model called DeepSeek-V4, which uses less computing power and can understand much longer texts than before. The company says this new model is advanced both in China and compared to other open-source AI systems globally.Key Facts
- DeepSeek first gained attention last year with a low-cost AI reasoning model similar to US versions.
- DeepSeek-V4 can handle a context of up to one million words, meaning it can process very long texts.
- The new model reduces computing and memory costs, making it cheaper and faster to use.
- There are two versions: V4-Pro with 1.6 trillion parameters and V4-Flash with 284 billion parameters; parameters help the model make decisions.
- DeepSeek-V4 is optimized for popular AI assistant tools like Claude Code and OpenCode.
- The V4-Pro model outperforms most open-source AI models and nearly matches Google’s top closed-source model, Gemini-Pro-3.1.
- DeepSeek’s AI has been adopted across China in healthcare, finance, and local government.
- The White House has accused Chinese companies of trying to steal US AI technology ahead of a planned summit between President Trump and Chinese leader Xi Jinping.
Read the Full Article
This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.