AI 모델
MimiMimi is an intelligent robot created by Miniwork that can answer all your questions quickly and accurately. Mimi will also continue to learn and update its knowledge to better meet your needs and become your right-hand man in study, work and life.
GPT-4.1Using OpenAI's most powerful model, GPT-4.1, it delivers more natural, engaging, and personalized writing while providing overall more comprehensive and insightful responses. It outperforms GPT-3.5 in quantitative tasks (mathematics and physics), creative writing, and many other challenging tasks. It supports a context window of 128k tokens.
OpenAI o3Powered by OpenAI's o3 model, it prioritizes reasoning capabilities and excels at complex tasks in science, programming, and mathematics. It also supports image reasoning and features a 200K token context window.
GPT-4o miniUsing OpenAI's most powerful model, GPT-4o mini, it delivers more natural, engaging, and personalized writing while providing overall more comprehensive and insightful responses.
Claude 4 SonnetClaude Sonnet 4 is a major upgrade of Claude Sonnet 3.7, which shows superior performance in both code writing and reasoning, and can respond to instructions more accurately.
Claude 3.7 SonnetAnthropic's advanced model version, with excellent performance. Excellent at solving complex problems, especially coding tasks.
Claude 3.5 HaikuAnthropic's latest model, ideal for code completion and interactive chatbots to enhance user-facing applications.
Gemini 2.5 FlashGoogle's next-generation thinking-enhanced model, comprehensively upgraded in thought process and reasoning demonstration, supports more transparent thinking processes and deeper logical analysis.
Gemini 2.5 ProGoogle's next-generation thinking-enhanced model, comprehensively upgraded in thought process and reasoning demonstration, supports more transparent thinking processes and deeper logical analysis.
DeepSeek-R1-0528DeepSeek DeepThink model delivers performance on par with OpenAI's o1 official release, excelling particularly in mathematics, coding, and natural language reasoning. Its Chinese processing capability is outstanding, achieving an accuracy rate of 92.7% in Chinese contexts—far surpassing ChatGPT.
Qwen-maxQwen is a general-purpose large language model series developed by Alibaba, supporting multilingual processing, code generation, and complex reasoning. It offers both open-source and commercial versions. Known for its robust Chinese language capabilities and long-context comprehension, Qwen is well-suited for enterprise services, developer tools, and AI research applications.
DoubaoDoubao is ByteDance's intelligent AI conversational assistant, equipped with natural language processing and multi-turn dialogue capabilities. It supports a variety of functions, including text generation, knowledge-based Q&A, code writing, creative writing, and study assistance. Additionally, it engages in personalized interactions by leveraging contextual understanding, making it suitable for work, study, and entertainment scenarios—delivering an efficient and seamless AI-powered experience for users.