电子行业行业动态跟踪点评：DeepSeek开源模型性价比突出，R1模型性能对标OpenAI o1正式版

Investment Rating - The industry investment rating is "Positive" as it is expected to outperform the benchmark index by more than 5% in the next six months [2][25]. Core Insights - DeepSeek's open-source model, DeepSeek-R1, offers a competitive price-performance ratio, matching the performance of OpenAI's o1 official version in various tasks such as mathematics, coding, and natural language reasoning [4]. - The pricing for DeepSeek-R1 API services is significantly lower than that of OpenAI, with costs of 1 RMB per million input tokens (cache hit) and 4 RMB (cache miss), compared to OpenAI's 55 RMB and 110 RMB respectively [4]. - Major tech companies like NVIDIA, Amazon, and Microsoft have integrated DeepSeek-R1 into their platforms, indicating strong market interest and adoption [5]. - The trend towards optimizing technology and commercial applications is evident, as companies are increasingly evaluating the efficiency of AI infrastructure investments [6]. - The introduction of distilled models by DeepSeek is expected to accelerate the deployment of AI at the edge, promoting local deployment of lightweight and cost-effective models [7]. Summary by Sections Industry Dynamics - On January 20, 2025, DeepSeek officially released DeepSeek-R1 and open-sourced its model weights, showcasing its competitive capabilities against OpenAI's offerings [3][4]. Market Adoption - By January 31, 2025, major U.S. tech giants announced their integration of DeepSeek-R1, with NVIDIA, Amazon, and Microsoft all incorporating the model into their services [5]. Technological Trends - The AI industry is shifting towards more efficient software models to reduce reliance on hardware like GPUs, as highlighted by TrendForce's insights on DeepSeek's recent model releases [6]. Investment Recommendations - The DeepSeek R1 model is seen as a significant breakthrough in terms of scale, efficiency, and cost, with potential beneficiaries including various semiconductor and AI-related companies [7].