Core Insights - OpenSeeker, developed by a research team from Shanghai Jiao Tong University, is the first fully open-source deep search agent with complete training data, breaking the data monopoly held by large companies [2][28]. - The model demonstrates that high-quality data synthesis can achieve state-of-the-art (SOTA) performance without relying on extensive computational resources [2][28]. Group 1: Model Development - OpenSeeker utilizes a unique high-quality data synthesis approach to overcome the data bottleneck typically faced by large enterprises [6][28]. - The model requires only 11.7k synthetic samples for a single round of supervised fine-tuning (SFT) to achieve competitive results on various benchmarks [17][28]. Group 2: Training Methodology - The training of deep search agents hinges on two critical aspects: creating challenging question-answer tasks and generating high-quality solution trajectories [7][8]. - OpenSeeker employs a fact-based question construction method using real web structures to ensure the model engages in genuine multi-hop reasoning [9][10][11]. - A dynamic denoising trajectory synthesis method is introduced to enhance core information extraction in noisy environments [12][15]. Group 3: Performance Metrics - OpenSeeker achieved a score of 48.4% on the BrowseComp-ZH leaderboard, surpassing Alibaba's Tongyi DeepResearch, which scored 46.7% after extensive training [17][18]. - The model's performance across multiple benchmarks includes 29.5 on BrowseComp, 48.4 on BrowseComp-ZH, 74.0 on xbench, and 59.4 on WideSearch [18]. Group 4: Data Quality and Challenges - The synthetic data generated by OpenSeeker presents a significantly higher difficulty level compared to existing benchmarks, with an average of 46.35 tool calls per trajectory and an average token length of 76.1k [25][20]. - In controlled data volume comparisons, OpenSeeker's data quality is notably superior to that of Alibaba's models, maintaining a significant advantage across various metrics [20][21]. Group 5: Community Impact - The open-source release of OpenSeeker is seen as a pivotal moment for advancing the field, providing researchers with a solid foundation for exploring next-generation search agents [24][28]. - The community response highlights the importance of data transparency and the ability to innovate without the constraints of data gatekeeping [26][29].

Investment Rating - The report maintains a "Buy" rating for the following companies: 罗博特科 (300757.SZ), 唯科科技 (301196.SZ), 能科科技 (603859.SH), and 合合信息 (688615.SH) [9][62]. Core Insights - The REDSearcher framework has achieved a breakthrough in deep search agent technology, overcoming three major industry bottlenecks and setting a new standard in the field [3][20][29]. - OpenAI's GPT-5.4 has been released, marking a new era for intelligent agents with its native computer operation capabilities, significantly enhancing its performance across various tasks [4][33][34]. - The AI financing landscape remains active, with significant investments in the CPO sector and brain-computer interface companies, indicating strong market confidence [48][51]. Summary by Sections Computing Power Dynamics - The REDSearcher team launched a low-cost, scalable deep search agent training framework, achieving state-of-the-art results with a 30B parameter model, surpassing several well-known closed-source models [3][20][29]. - The rental prices for computing power remain stable, with specific pricing details for various configurations provided [20][21]. AI Application Dynamics - Bing's weekly traffic increased by 2.70%, while ChatGPT and Gemini also showed significant traffic figures, indicating a competitive landscape in AI applications [31][32]. - GPT-5.4 is the first general model to possess native computer operation capabilities, allowing it to perform complex tasks typically handled by humans [33][34]. AI Financing Trends - AyarLabs completed a $500 million E round financing, reaching a valuation of $3.75 billion, while NVIDIA invested $2 billion each in Lumentum and Coherent to enhance its AI infrastructure [48][50]. - Science Corporation raised $230 million in C round financing, achieving a valuation of $1.5 billion, focusing on the commercialization of its retinal implant technology [51][52]. Investment Recommendations - Broadcom reported a strong Q1 2026 performance with revenues of $19.311 billion, driven by a 106% year-on-year growth in AI business revenue, reinforcing the positive outlook for AI infrastructure investments [5][60]. - The report suggests focusing on companies with high growth potential in the AI sector, including 唯科科技, 合合信息, 能科科技, and 罗博特科 [61].