Chinese AI startup DeepSeek has launched an upgraded version of its artificial intelligence model, DeepSeek-V3-0324, which boasts enhanced programming capabilities and improved performance metrics. The new model is designed to compete with leading AI systems from OpenAI, Google, and others. Reports indicate that DeepSeek-V3 is outperforming models such as GPT-4.5, Claude 3.7 Sonnet, and Gemini 2.0 in various benchmarks. Notably, it achieves speeds of over 250 tokens per second and has demonstrated a significant increase in reasoning and front-end skills. The upgrade has prompted a wave of low-cost AI offerings from Chinese tech firms, challenging the premium pricing of competitors. Analysts suggest that DeepSeek's advancements could shift the competitive landscape in the AI sector, particularly as it utilizes inexpensive chips to deliver high performance. The model is now available on platforms like Hugging Face and Together AI, with tests indicating it is three times faster than its predecessor, V2.