Chatvim

AI chat with markdown files in Neovim.

Unlike many other Neovim AI plugins, Chatvim uses a plain markdown document as the chat window. No special dialogs or UI elements are required.

Video Demo

https://github.com/user-attachments/assets/7fe523e8-b953-4307-ac62-df58884bdfd0

Usage

Simply type a message into a file in Neovim, and then run the :ChatvimComplete command to get a response from the default AI model. With no extra configuration, the entire file is treaged as the user prompt. See below for more details.

Features

The purpose of Chatvim is to treat a markdown document as a conversation with an AI assistant instead of using a separate chat window or dialog box.

You can save, copy, fork, version, and share the markdown document as you would with any other markdown file. This is easier for some workflows than using a separate chat interface.

Because Chatvim uses chat completion AI models, there must be some way to separate user messages, assistant messages, and system messages. This is done using delimiters in the markdown document.

The default delimiters are:

• # === USER === for user messages
• # === ASSISTANT === for assistant messages
• # === SYSTEM === for system messages

You can customize these delimiters in the markdown front matter, explained below.

If no delimiter is used (as is the case for most markdown documents), then the entire markdown document is treated as user input, and the AI model will respond to it as if it were a single user message. Delimiters will be added with the first response.

Installation

You must install node.js v24+.

You must also set at least one API key:

• Set XAI_API_KEY environment variable to your xAI API key.
• Set OPENAI_API_KEY environment variable to your OpenAI API key.

Then, install with LazyVim:

{
  "chatvim/chatvim.nvim",
  build = "npm install",
  config = function()
    require("chatvim")
  end,
},

Or with Packer:

use {
  "chatvim/chatvim.nvim",
  run = "npm install",
  config = function()
    require("chatvim")
  end,
}

Commands

:ChatvimComplete

Completes the current markdown document using the AI model. If no delimiters are present, it will treat the input as user input and append a response.

:ChatvimStop

Stops the current streaming response.

:ChatvimNew [direction]

direction can be blank or left, right, top, or bottom. Opens a new (unsaved) markdown document in a new split window for a new chat session. If direction is not specified, it defaults to the current window.

Recommended Shortcuts

Add these shortcuts to your nvim configuration to make it easier to use Chatvim.

local opts = { noremap = true, silent = true }
vim.api.nvim_set_keymap("n", "<Leader>cvc", ":ChatvimComplete<CR>", opts)
vim.api.nvim_set_keymap("n", "<Leader>cvs", ":ChatvimStop<CR>", opts)
vim.api.nvim_set_keymap("n", "<Leader>cvnn", ":ChatvimNew<CR>", opts)
vim.api.nvim_set_keymap("n", "<Leader>cvnl", ":ChatvimNew left<CR>", opts)
vim.api.nvim_set_keymap("n", "<Leader>cvnr", ":ChatvimNew right<CR>", opts)
vim.api.nvim_set_keymap("n", "<Leader>cvnb", ":ChatvimNew bottom<CR>", opts)
vim.api.nvim_set_keymap("n", "<Leader>cvnt", ":ChatvimNew top<CR>", opts)

Configuration

Use "+++" for TOML or "---" for YAML front matter. Front matter is used to specify settings for Chatvim. Place the front matter at the top of your markdown document, before any content. The front matter should look like this:

+++
model = "grok-3"  # or "gpt-4.1", etc.
delimiterPrefix = "\n\n"
delimiterSuffix = "\n\n"
userDelimiter = "# === USER ==="
assistantDelimiter = "# === ASSISTANT ==="
systemDelimiter = "# === SYSTEM ==="
+++

All fields are optional.

Models

Models supported:

• claude-sonnet-4-0 (Anthropic)
• claude-opus-4-0 (Anthropic)
• claude-3-7-sonnet-latest (Anthropic)
• claude-3-5-sonnet-latest (Anthropic)
• gpt-4.1 (OpenAI)
• gpt-4.1-mini (OpenAI)
• gpt-4.1-nano (OpenAI)
• gpt-4o (OpenAI)
• gpt-4o-mini (OpenAI)
• gpt-4o-mini-search-preview (OpenAI)
• gpt-4o-search-preview (OpenAI)
• o3 (OpenAI)
• o3-mini (OpenAI)
• o1 (OpenAI)
• o1-mini (OpenAI)
• grok-3 (default) (xAI)
• grok-4-0709 (xAI)
• grok-3-mini (xAI)
• grok-3-fast (xAI)
• grok-3-mini-fast (xAI)

Providers supported:

• xAI (Set XAI_API_KEY environment variable to your xAI API key)
• OpenAI (Set OPENAI_API_KEY environment variable to your OpenAI API key)
• Anthropic (Set ANTHROPIC_API_KEY environment variable to your Anthropic API key)

License

MIT license. See LICENSE.

Copyright (C) 2025 Ryan X. Charles