AI产业跟踪：openAI发布GPT-4.1系列，智能体应用有望加速落地

Investment Rating - The report maintains a "Positive" investment rating for the software and services industry [8]. Core Insights - On April 15, 2025, OpenAI released the GPT-4.1 series, which includes models GPT-4.1, GPT-4.1 Mini, and GPT-4.1 Nano, available through API calls to all developers. The series offers superior performance at a lower cost compared to previous models, with a planned discontinuation of the GPT-4.5 preview version on July 14, 2025 [2][5]. - The GPT-4.1 series supports a context of 1 million tokens, enhancing its applicability in complex scenarios such as finance and law. The flagship model, GPT-4.1, shows significant improvements in programming and writing capabilities, while the Mini and Nano models offer cost-effective solutions for mid-range and edge applications [10]. - The pricing for GPT-4.1 has been reduced, with input and output costs set at $2 and $8 per million tokens, respectively, which is 26% cheaper than GPT-4o. The Nano model is priced even lower, making it competitive in the B-end market [10]. Summary by Sections Event Description - OpenAI's GPT-4.1 series was launched, showcasing performance comparable to or better than the GPT-4.5 preview version, with a focus on cost-effectiveness [5]. Event Commentary - The report emphasizes the potential for the GPT-4.1 series to drive advancements in the B-end market, particularly through partnerships with AI agent companies, domestic AI chip leaders like Cambricon, cloud service providers collaborating with DeepSeek, and IDC firms working with major players like Tencent and Alibaba [10].