Skip to content

Get Started

DeepTutor ships four installation paths. They all share one workspace layout: settings live in data/user/settings/ under the directory you launch from (or under DEEPTUTOR_HOME / deeptutor start --home if you set one explicitly). Knowledge bases, memory, Partner runtime state, multi-user accounts, and audit logs also live under the same data/ tree, so deployments only need to persist that tree.

For the full app, the recommended flow is pick a workspace directory → install → deeptutor initdeeptutor start.

OptionWhen to usePage
PyPIFirst-time users, want the smoothest path, no cloneInstall from PyPI
From SourceHacking on the codebase, want hot reloadInstall from Source
DockerProduction deployments, hosted hosts, persistent volumesDocker
CLI OnlyHeadless server, agent harness, no Web UI neededCLI-Only Install

🧪 Trying a pre-release? Plain pip install -U deeptutor stays on stable — PyPI skips pre-release versions. Opt in with pip install --pre -U deeptutor, or pin an exact version with pip install -U "deeptutor==<version>".

ComponentVersion
Python3.11–3.13
Node.js20+ (PyPI) / 22 LTS (from source, matches CI + Docker)
RAM4 GB free
Disk2 GB free

You’ll also need an LLM API key (OpenAI, Anthropic, Gemini, DeepSeek, Azure, Ollama, LM Studio, llama.cpp, vLLM, or any OpenAI-compatible endpoint) and ideally an embedding API key if you plan to use Knowledge Bases.

After deeptutor start, two services come up:

ServiceDefault port
FastAPI backend8001
Next.js frontend3782

Both are overridable via data/user/settings/system.json or during deeptutor init.

Once you can reach http://localhost:3782/:

If something breaks, jump to Troubleshooting.