NEKOVA docs
NEKOVA / Docs / Getting Started

Installation

NEKOVA installs like any other Python package — one command, no separate runtime to manage.

Requirements

NEKOVA runs on top of Python 3.10 or later. You don't need to know Python to write NEKOVA programs, but the interpreter itself is a Python package, so pip is how it gets onto your machine.

Install from PyPI

pip install nekova-lang

This installs the nekova command-line tool, which gives you everything: running files, the REPL, the project scaffolder, and the built-in checker.

Verify it worked:

nekova --version

Your first file

Create a file called hello.nk:

show "Hello from NEKOVA"

Run it:

nekova run hello.nk

The AI provider

NEKOVA ships with a mock AI provider enabled by default, so think and friends work immediately with no API key — every mock response is tagged [MOCK] so it's never mistaken for a real model reply. This is deliberate: you shouldn't need to sign up for anything to try the language.

When you're ready to use a real model, set the relevant API key as an environment variable (the exact variable name depends on which provider you configure) and NEKOVA will use it automatically. See think for details.

VS Code extension

If you use VS Code, search the marketplace for "NEKOVA - AI Native Programming Language" (publisher: SYNEKCOTTech), or install directly:

code --install-extension SYNEKCOTTech.nekova-lang

It adds syntax highlighting, snippets, two color themes built from NEKOVA's own brand palette, and a .nk file icon.

Next
Head to the Quickstart for a guided tour of the language, or jump straight to Core Syntax if you'd rather explore on your own.