Teach anything.
Build open-access, course-specific AI chatbots using open-source LLMs for your students. Upload materials and customize responses—all for free.
Trusted By Professors At
Upload. Chat. Learn.
Use open-source, non-commercial large language models (LLMs) to transform your course materials into an open-access, multilingual chat experience.
Open Access
Designed specifically for professors and students in higher education.
Privacy Focused
Your data stays yours. We don't use it to train any LLM.
Open Source
We only use open-source LLMs. Fully transparent code. Community-driven development.
Always free. Always open.
No hidden costs, no premium tiers, no vendor lock-in.
Built on modern infrastructure.
Powered by cutting-edge open-source LLMs and retrieval-augmented generation. Your chatbots understand context and provide accurate, source-based answers.
RAG-Powered Accuracy
Responses grounded in your uploaded materials
Multiple open-source LLMs
Choose from leading open-source language models
Fast & Reliable
Built on modern serverless infrastructure that is scalable and efficient.
How It Works
Create
Sign up and create your first chatbot in minutes. No technical skills needed. Select from open-source LLMs such as Mistral, Llama 3.3, and Qwen.
Upload
Add your course materials. Our AI processes and indexes everything automatically.
Share
Share the link with your students and watch engagement soar. Or embed the chatbot on your own website with automatically-generated HTML or Javascript code.
Why Choose AI Alexa?
See how we compare to commercial AI solutions. Built with privacy, openness, and educators in mind.
| Feature | AI Alexa (open access) | Commercial AI |
|---|---|---|
| Audience | For educators | General purpose |
| Access to custom chatbot | Shareable open-access link to custom chatbot, no login required | User account required to see shared chatbots |
| Student login requirement | Students do not need to login | Students must create accounts |
| Privacy and data protection | Full privacy protection; files & chats not used to train LLMs | User data used for AI training |
| Openness of platform | Open source, open access | Free tier product is not open access; training data not transparent |
| Model choice | Choice among open-source LLMs | Locked in to one company's LLM |
| Control of system prompts | Professors write system prompts | Prompts controlled by provider |
| Role of faculty and students | Faculty & students as designers, not just consumers | Merely consumers of a predefined product |
| Building custom chatbots from files | File uploads to create custom chatbots; files remain private | File uploads available but files are used by AI companies for LLM training |
| Data usage for model training | Files & chats not used to train LLMs | Content is used to train LLM |
| Language support | Multilingual | Multilingual |
Our commitment to open access, privacy, and educator empowerment sets us apart from commercial AI solutions that prioritize data collection and vendor lock-in.
Empowering Professors with Open-Access AI
Built by Akhilesh Rangani under the guidance of Professor Alexa Alice Joubin at Digital Humanities Institute at George Washington University.
Professor Alexa Alice Joubin at the QS Higher Ed Summit in Washington, D.C., June 4, 2024
Open source.
Forever free.
Built by educators for educators. Contribute to best practices. Share pedagogical ideas. Collaborate on GitHub. Use our hosted AI or host it yourself.
No credit card required • Academic emails preferred