[Rakuten Group Opened Japanese Large Language Model Rakuten AI 3.0, Allegedly Based on DeepSeek V3 Architecture, Sparking Controversy] Rakuten Group announced the launch of its Japanese-specialized large language model, Rakuten AI 3.0. The company claims that the model outperforms models like GPT-4o in various Japanese benchmarks, including Japanese culture and history, instruction adherence, and more. It was developed by combining open-source community models with proprietary bilingual data. However, users on X discovered that the config.json file on the project's Hugging Face page contains entries like `"model_type": "deepseek_v3"`. Additionally, when the model responds to related questions, its stance on public opinion appears to lean more toward China rather than Japan. This has sparked discussions among users about whether the model was developed based on DeepSeek and how much of the technology is independently developed in Japan.