DeepSeek Upgrades AI Model to Rival ChatGPT, Gemini
18
1
Chinese artificial intelligence company DeepSeek has unveiled a major upgrade to its chatbot, enhancing logic, mathematics, and programming capabilities while reducing hallucinated responses. The new model, DeepSeek-R1-0528, now rivals OpenAI’s GPT-4 and Google’s Gemini 2.5 Pro in reasoning and inference tasks.
DeepSeek’s January debut of its R1 chatbot marked China’s rising influence in AI, delivering top-tier performance at a fraction of the cost—just $6 million compared to competitors’ massive training budgets. The latest iteration boosts accuracy on complex tests like AIME 2025 from 70% to 87.5%, processing nearly double the data per query for deeper analysis.
By:
Previous
SEC Postpones XRP ETF Decision as CME Exchange Sees Strong XRP Futures Debut
Next
Pi Network Price Plummets Below Critical Support Amid Bearish Signals
