curiously.app

What It Is

Curiously is an AI application built for educational settings, positioned as a tool that helps educators create custom AI teaching assistants and learning companions. Based on the scraped text, it emphasizes combining large language model research with best practices, and uses retrieval-augmented generation (RAG) methods so teachers can create an AI companion better aligned with their own course materials and classroom needs.

Core Capabilities and Use Cases

Its core capability is a “custom AI teaching assistant / learning companion.” In practice, this means teachers may be able to configure an AI assistant around course content, class needs, or learning objectives, and use it for student Q&A, learning support, and answering questions about course materials. The text explicitly mentions RAG, which usually means the system attempts to combine external materials or course knowledge with large-model generation to improve the relevance of answers to the teaching content. However, the page does not disclose which types of materials are supported, whether source citations are provided, whether multi-turn learning records are supported, or what kind of underlying model it uses.

Pricing, Chinese Support, and Integrations

The currently scraped information does not provide details on a free tier, trial policy, paid plans, or payment methods, so its value for money can only be assessed conservatively. Chinese-language support is also not mentioned, making it unclear whether it is suitable for Chinese-language classrooms or bilingual teaching. There is likewise no public information about API access, LMS integrations, or compatibility with education systems such as Google Classroom or Canvas. For schools or institutional procurement, these missing details will make evaluation less efficient.

Pros, Cons, and Limitations

Its main advantage is its vertical focus: it is designed around helping teachers build AI teaching assistants, making it more closely aligned with education workflows than general-purpose chatbots. It also mentions LLMs and RAG, which fits the need for course-material-based Q&A. The limitation is that there is very little public information available: there are no case studies, output examples, privacy policy summary, student data handling explanation, or descriptions of safety guardrails, hallucination control, and teacher review mechanisms. Educational use cases have high requirements for accuracy and data compliance, so further testing is essential before real-world deployment.

Who It’s For and Access from China

Curiously is suitable for teachers, instructional designers, and education institutions that want to explore AI teaching assistants, course Q&A assistants, or personalized learning companions. Access from mainland China is currently unknown, and network connectivity, payment methods, and the Chinese-language experience all need to be tested in practice. If access or compliance is limited, alternatives include Khanmigo, MagicSchool AI, Eduaide.AI, ChatGPT custom GPTs, or building a course knowledge-base Q&A bot using domestic large-model platforms.