An independent analysis company for models and APIs, Artificial Analysis, released its latest evaluation report on DeepSeek R1-0528. The results show that DeepSeek R1 has made significant breakthroughs in technical performance, surpassing xAI, Meta, and Anthropic all at once, standing shoulder to shoulder with Google as the world's second-largest artificial intelligence laboratory. At the same time, DeepSeek has established an undisputed leadership position in the field of open-source weights.
According to Artificial Analysis's Intelligence Index, the DeepSeek R10528 version jumped from 60 points to 68 points in seven mainstream evaluations of all top-tier models.
This improvement is equivalent to the change between OpenAI's o1 model and o3 model (from 62 points to 70 points). This significant progress not only demonstrates the strong capabilities of DeepSeek R1 in the field of artificial intelligence but also lays a foundation for future development.
Key points:
🌟 DeepSeek R1-0528 surpassed xAI, Meta, and Anthropic, becoming the world's second-largest AI laboratory.
📈 The Intelligence Index score jumped from 60 points to 68 points, with an improvement comparable to OpenAI's o1 to o3 models.
🚀 DeepSeek has established an undisputed leadership position in the field of open-source weights, promoting the popularization and innovation of technology.
