Skip to main content
DeepTutor has listed Eden AI as a built-in provider since v1.5.5, so you pick it from a list in the setup wizard instead of configuring a custom endpoint. This page wires Eden AI as DeepTutor’s LLM provider only.

​
Overview

DeepTutor is a self-hosted AI study and research app with chat, quizzes and knowledge bases. Eden AI is registered as a built-in gateway provider in its model registry, so the instance can run on models from OpenAI, Anthropic, Google, Mistral and more behind one key. This page configures the LLM only. DeepTutor has no Eden AI embedding binding, so the embedding step for knowledge bases is not covered here.

​
Prerequisites

  • Python 3.11 to 3.14 and Node.js 20 or newer on your PATH
  • An Eden AI API key from app.edenai.run

​
Setup

Install DeepTutor in a new directory and run the setup wizard:
In the wizard’s LLM step:
  1. Choose Show all providers, then Eden AI. Eden AI is not in the short featured list.
  2. Accept the default base URL, https://api.edenai.run/v3.
  3. Paste your Eden AI API key. If EDENAI_API_KEY is already exported in your shell, the wizard offers to reuse it.
  4. Pick a model from the list the wizard fetches from <base URL>/models. If that request fails, the wizard asks you to type the model name.
Open the frontend URL printed in the terminal (default http://127.0.0.1:3782). Already running an instance? Change the provider later in the web UI under Settings, in Providers and Language models. DeepTutor recognises Eden AI from the edenai keyword in the base URL.

​
Where the key is stored

DeepTutor saves the profile in data/user/settings/model_catalog.json: DeepTutor does not read a project .env file. Put the key in the wizard or the Settings UI, or export EDENAI_API_KEY in your shell before deeptutor init.

​
Model naming

Use the provider/model format, for example anthropic/claude-sonnet-latest, openai/gpt-latest or mistral/mistral-large-latest. Browse the catalog in List LLM Models.

​
Troubleshooting

​
The wizard asks for a model name

The wizard fetches models from <base URL>/models, and when that request fails it prompts for a custom model name. Type a provider/model id from the catalog, or check the base URL and your network and re-run deeptutor init.

​
The key in .env has no effect

DeepTutor ignores project-root .env files. Re-enter the key in Settings or export EDENAI_API_KEY in the shell that runs deeptutor init.

​
Next Steps