> Graphify: the code knowledge graph for agents. Canonical: https://graphify.com/blog/graphifyy-vs-graphify-on-pypi-which-package-to-install · Full index: https://graphify.com/llms-full.txt

# graphifyy pip package vs graphify: which to install

*April 22, 2026 · Syed Fahad*

The Graphify CLI is published on PyPI as `graphifyy`, spelled with two y's. Install it with `uv tool install graphifyy`, `pipx install graphifyy`, or `pip install graphifyy`. The command you run afterwards is still `graphify`. A PyPI package with the single-y name is a different project and is not affiliated with Graphify.

## Which package do you install with pip: graphifyy or graphify?

Install `graphifyy`. That is the only official package for the open-source Graphify CLI. The project README says so in an "Official package" note. It also says other packages whose names start with `graphify` are not affiliated. If you typed the single-y name, uninstall it and install `graphifyy`.

| | `graphifyy` | Single-y name on PyPI |
|---|---|---|
| Project | Open-source Graphify CLI | Unrelated project |
| Command it installs | `graphify` | Not the Graphify CLI |
| Source | [Graphify on GitHub](https://github.com/Graphify-Labs/graphify) | Not this repository |
| PyPI page | [pypi.org/project/graphifyy](https://pypi.org/project/graphifyy/) | Not this project |

## Why is the package called graphifyy?

The single-y name belongs to a different project on PyPI. PyPI compares project names after normalizing them, as the [Python Packaging User Guide](https://packaging.python.org/en/latest/specifications/name-normalization/) explains, and each normalized name maps to one project. So the Graphify package carries a second y, and the command keeps the short name.

The split is common in Python. The name you install and the command you run do not have to match.

## How do you install graphifyy?

Graphify needs Python 3.10 or newer. The recommended way is a tool installer such as uv or pipx. A tool installer puts the CLI in its own isolated environment and adds the `graphify` command to your PATH. That keeps its dependencies away from your project's own packages and makes upgrades a single command:

```bash
uv tool install graphifyy
graphify install
```

`uv tool install` comes from [uv](https://docs.astral.sh/uv/guides/tools/). `pipx install graphifyy` does the same job if you use [pipx](https://pipx.pypa.io/). Plain `pip install graphifyy` works too, inside a virtual environment.

`graphify install` copies the Graphify skill into your coding assistant. For an assistant other than Claude Code, pass `--platform`, for example `graphify install --platform codex`.

## How do you add optional extras?

Some features need extra dependencies that the base package leaves out to stay small. Examples include reading PDFs, Office files, or video, or connecting to a database. Install them by adding the extra name in square brackets after `graphifyy`. Quote the whole name so your shell does not try to expand the brackets:

```bash
uv tool install "graphifyy[mcp]"     # MCP server
uv tool install "graphifyy[office]"  # .docx and .xlsx
uv tool install "graphifyy[video]"   # local audio and video transcription
```

## How do you check you have the right package?

Run `pip show graphifyy` or `uv tool list`. You should see `graphifyy` and its version number. Then run `graphify --help`. The help text lists commands such as `install`, `query`, `path`, `explain`, and `benchmark`. If you see none of these, you may have installed the unrelated single-y package by mistake.

If `graphify` is not found after a pip install, your scripts folder is not on PATH. The README lists the common locations. A tool installer avoids this, and so does `python -m graphify`.

## Why does this matter for security?

Installing a package runs code from its author. A near-miss name is the classic setup for typosquatting, where a look-alike package catches mistyped installs. Check the package name, the PyPI project links, and the repository before you install any tool. Pin versions in CI.

## What if the graphify command is not found?

The package installed, but its folder is not on your PATH. With uv, run `uv tool update-shell` and open a new terminal. With pipx, run `pipx ensurepath`. With plain pip, add your Python scripts folder to PATH yourself. The command is always `graphify`, even though the package is `graphifyy`.

Check with `graphify --help`. If it prints the command list, the install worked.

## FAQ

### Is graphifyy the official Graphify package?

Yes. The README and the PyPI project page both point to the Graphify repository.

### Why does the command say graphify, not graphifyy?

The package name and the command name are separate. The package declares a `graphify` command when it installs.

### Can I install graphifyy with uv?

Yes. `uv tool install graphifyy` installs it as an isolated tool and adds it to PATH.

### What Python version does graphifyy need?

Python 3.10 or newer.

### I installed the wrong package. What now?

Uninstall the single-y package, then install `graphifyy`. Run `graphify --help` to confirm.

## Next steps

The code is on [GitHub](https://github.com/Graphify-Labs/graphify). Install with `uv tool install graphifyy` and follow the [installation docs](https://docs.graphify.com/open-source/installation). Once installed, read [what an MCP server is](/blog/what-is-an-mcp-server-a-developers-guide) to serve your graph to an agent. The [Claude Code integration page](https://graphify.com/integrations/claude-code) covers what to do next. See also [measuring token spend per task](/blog/measuring-token-spend-per-task-in-claude-code).