Tutor-GPT

Tutor-GPT is a self-hosted browser application for learners who want an AI tutor that adjusts to their learning needs. You can host its Next.js server yourself, including through Docker. Model responses come from a separate inference provider, with OpenRouter integration sending requests to an external service.

Developed by Plastic Labs, the tutor uses theory-of-mind reasoning to infer a learner's needs and revise the prompts behind its replies. Honcho builds a representation of each user that informs personalization. The tutoring scope extends beyond literacy.

The application accepts a chosen model through an OpenAI-compatible API endpoint. Hosting the web application gives you control over that application, while the configured inference provider processes its model requests. Honcho handles learner modeling through a separate service connection.

Supabase manages user accounts and database storage. It can run locally through Docker, so those parts can stay on your own infrastructure. The application also connects to Mistral OCR, and optional Sentry monitoring and PostHog analytics introduce additional service connections when enabled.

Tutor-GPT is open source under GPL-3.0, so operators can modify their own tutoring application under that license. Bloom is the hosted version of Tutor-GPT.