codegpt-ng is a minimalist AI plugin for neovim centered around a command based workflow. It has full support for Ollama, OpenAI, Azure, Anthropic and Groc APIs. You can easilly define new commands, custom prompts with a template system, and model configurations without Lua code.
This is a fork of the original CodeGPT repository from github user @dpayne. All credit goes to him for the initial work.
This fork does the following:
- Full support for Ollama and OpenAI API
- Streaming mode for real-time popup responses
- New table-based configuration instead of global variables
- New commands and added support to the
%range modifier - Ability to cancel current request.
- UI Query and select local or remote model
- Strips thinking tokens from replies if the model forgets to use codeblocks
- New callback types:
insert_linesandprepend_lines - Model definition inheritance: Define models that inherit other model parameters
- Refactored for idiomatic Lua and neovim plugin style
- Simplified command definition with explicit configuration specification
- Chat History: Add example messages in a command definition
- Tests with plenary library
- Fixed statusline integration
Although this fork introduces breaking changes and a substantial rewrite, I've tried to preserve the original project's minimalist spirit — a tool that connects to AI APIs without getting in the way. The goal remains to provide simple, code-focused interactions that stay lightweight and unobtrusive, letting developers leverage LLMs while maintaining control over their workflow.
In particular, the model definition flow was carefully designed to quickly add custom model profiles for specific cases and easily switch between them or assign them to custom commands.
- The plugins 'plenary' and 'nui' are also required.
Installing with Lazy.
{
"blob42/codegpt-ng.nvim",
dependencies = {
'nvim-lua/plenary.nvim',
'MunifTanjim/nui.nvim',
},
opts = {
-- configuration here
}
}Installing with Packer.
use({
"blob42/codegpt-ng.nvim",
requires = {
"MunifTanjim/nui.nvim",
"nvim-lua/plenary.nvim",
},
config = function()
require("codegpt").setup({
-- Configuration here
})
end
})The top-level command is :Chat. The behavior is different depending on whether text is selected and/or arguments are passed.
:Chatwith text selection will trigger thecompletioncommand, ChatGPT will try to complete the selected code snippet.
chat_selection.mp4
:Chat some instructionswith text selection and command args will invoke thecode_editcommand.
code_edit.mp4
:Chat <command>: if there is only one argument and that argument matches a command, it will invoke that command with the given text selection.
Here are a few example commands to illustrate it:
:Chat docgenerates documentation blocks and prepends it to the selected text.- Use
codegpt.select_model()to quickly select different models withvim.ui.select
doc-cmd-select-model.mp4
In the below example :Chat tests will attempt to write units for the selected code.
tests-cmd.mp4
- Ask question about the selected text file. This demo also showcases using the
%range modifier to use all the buffer as selection.
percent-question-cmd.mp4
:Chat hello worldwithout any text selection will trigger thechatcommand.- Streaming can be toggled in the config
Note: you have to input at least two words otherwise it would be considered as a codegpt command.
chat-mode-streaming.mp4
- Range character
%Chatto select all buffer. :VChat: to temporary enforce the vertical layout.Chat!: To make popup window persistent when the cursor leaves.
Here is the full list of predefined command actions:
| command | input | Description |
|---|---|---|
| completion | text selection | Will ask ChatGPT to complete the selected code. |
| code_edit | text selection + command args | Will ask ChatGPT to apply the given instructions (the command args) to the selected code. |
| explain | text selection | Will ask ChatGPT to explain the selected code. |
| question | text selection | Will pass the commands args to ChatGPT and return the answer in a text popup. |
| debug | text selection | Will pass the code selection to ChatGPT analyze it for bugs, the results will be in a text popup. |
| doc | text selection | Will ask ChatGPT to document the selected code. |
| opt | text selection | Will ask ChatGPT to optimize the selected code. |
| tests | text selection | Will ask ChatGPT to write unit tests for the selected code. |
| chat | command args | Will pass the given command args to ChatGPT and return the response in a popup. |
require("codegpt").setup({
api = {
provider = "openai", -- or "Ollama", "Azure", etc.
openai_api_key = vim.fn.getenv("OPENAI_API_KEY"),
chat_completions_url = "https://api.openai.com/v1/chat/completions",
},
models {
-- model definitions
},
commands = {
-- Command defaults
},
ui = {
-- UI configuration
},
hooks = {
-- Status hooks
},
clear_visual_selection = true,
})The configuration table commands can be used to override command configurations.
commands = {
completion = {
model = "gpt-3.5-turbo",
user_message_template = "This is a template...",
callback_type = "replace_lines",
},
tests = {
language_instructions = { java = "Use TestNG framework" },
}
}Custom commands can be added to the commands table.
commands = {
modernize = {
user_message_template = "Modernize the code...",
language_instructions = { cpp = "..." }
}
}The models table defines available LLM models for each provider. Models are
organized by provider type and can inherit parameters from other models.
models = {
default = "gpt-3.5-turbo", -- Global default model
ollama = {
default = "gemma3:1b", -- Ollama default model
['qwen3:4b'] = {
alias = "qwen3", -- Alias to call this model
max_tokens = 8192,
temperature = 0.8,
append_string = '/no_thinking', -- Custom string to append to the prompt
},
},
openai = {
["gpt-3.5-turbo"] = {
alias = "gpt35",
max_tokens = 4096,
temperature = 0.8,
},
},
}Models can inherit parameters from other models using the from field. For example:
["gpt-foo"] = {
from = "gpt-3.5-turbo", -- Inherit from openai's default
temperature = 0.7, -- Override temperature
}Use alias to create shorthand names for models.
Specify model parameters like max_tokens, temperature, and append_string
to customize behavior. see lua/codegpt/config.lua file for the full config specification.
- Call
:lua codegpt.select_model()to interactively choose a model via UI.
ui = {
popup_type = "popup", -- or "horizontal", "vertical"
text_popup_filetype = "markdown",
mappings = {
quit = "q",
use_as_output = "<c-o>",
},
popup_options = {
relative = "editor",
position = "50%",
size = { width = "80%", height = "80%" },
},
popup_border = { style = "rounded" },
popup_window_options = { wrap = true, number = true },
}hooks = {
request_started = function() vim.cmd("hi StatusLine ctermfg=yellow") end,
request_finished = function() vim.cmd("hi StatusLine ctermfg=NONE") end,
}You can use macros inside the user/system message templates when defining a command.
The system_message_template and user_message_template can contain the following macros:
| macro | description |
|---|---|
{{filetype}} |
The filetype of the current buffer. |
{{text_selection}} |
The selected text in the current buffer. |
{{language}} |
The name of the programming language in the current buffer. |
{{command_args}} |
Everything passed to the command as an argument, joined with spaces. |
{{language_instructions}} |
The found value in the language_instructions map. |
Here is are a few examples to demonstrate how to use them:
commands = {
--- other commands
cli_helpgen = {
system_message_template = 'You are a documentation assistant to a software developer. Generate documentation for a CLI app using the --help flag style, including usage and options. Only output the help text and nothing else.',
user_message_template = 'App name and usage:\n\n```{{filetype}}\n{{text_selection}}```\n. {{command_args}}. . {{language_instructions}}',
model = 'codestral',
language_instructions = {
python = 'Use a standard --help flag style, including usage and options, with example usage if needed.',
},
},
rs_mod_doc = {
system_message_template = 'You are a Rust documentation assistant. Given the provided source code, add appropriate module-level documentation that goes at the top of the file. Use the `//!` comment format and example sections as necessary. Include explanations for what each function in the module.',
user_message_template = 'Source code:\n```{{filetype}}\n{{text_selection}}\n```\n. {{command_args}}. Generate the doc using module level rust comments `//!` ',
},
-- dummy command to showcase the use of chat_history
acronym = {
system_message_template = 'You are a helpful {{filetype}} programming assistant that abbreviates identifiers and variables..',
user_message_template = 'abbreviate ```{{filetype}}\n{{text_selection}}```\n {{command_args}}',
chat_history = {
{ role = 'user', content = 'abbreviate `configure_user_script`' },
{ role = 'assistant', content = 'c_u_s' },
{ role = 'user', content = 'abbreviate ```lua\nlocal = search_common_pattern = {}```\n' },
{ role = 'assistant', content = 'local = s_c_p = {}' },
},
},| name | Description |
|---|---|
| text_popup | Will display the result in a text popup window. |
| code_popup | Will display the results in a popup window with the filetype set to the filetype of the current buffer. |
| replace_lines | Replaces the current lines with the response. If no text is selected, it will insert the response at the cursor. |
| insert_lines | Inserts the response after the current cursor line without replacing any existing text. |
| prepend_lines | Inserts the response before the current lines. If no text is selected, it will insert the response at the beginning of the buffer. |
require("codegpt").setup({
-- Connection settings for API providers
connection = {
api_provider = "openai", -- Default API provider
openai_api_key = vim.fn.getenv("OPENAI_API_KEY"),
chat_completions_url = "https://api.openai.com/v1/chat/completions", -- Default OpenAI endpoint
ollama_base_url = "http://localhost:11434", -- Ollama base URL
proxy = nil, -- Can also be set with $http_proxy environment variable
allow_insecure = false, -- Disable insecure connections by default
},
-- UI configuration for popups
ui = {
stream_output = false, -- Disable streaming by default
popup_border = { style = "rounded", padding = { 0, 1 } }, -- Default border style
popup_options = nil, -- No additional popup options
text_popup_filetype = "markdown", -- Default filetype for text
popups
popup_type = "popup", -- Default popup type
horizontal_popup_size = "20%", -- Default horizontal size
vertical_popup_size = "20%", -- Default vertical size
spinners = { "", "", "", "", "", "" }, -- Default spinner icons
spinner_speed = 80, -- Default spinner speed
actions = {
quit = "q", -- Quit key
use_as_output = "<c-o>", -- Use as output key
use_as_input = "<c-i>", -- Use as input key
cancel = "<c-c>", -- cancel current request
custom = nil, -- table. with custom actions
},
},
-- Model configurations grouped by provider
models = {
default = "gpt-3.5-turbo", -- Global default model
ollama = {
default = "gemma3:1b", -- Ollama default model
['qwen3:4b'] = {
alias = "qwen3", -- Alias to call this model
max_tokens = 8192,
temperature = 0.8,
append_string = '/no_think', -- Custom string to append to the prompt
},
},
openai = {
["gpt-3.5-turbo"] = {
alias = "gpt35",
max_tokens = 4096,
temperature = 0.8,
},
},
},
-- General options
clear_visual_selection = true, -- Clear visual selection when the command starts
-- Custom hooks
hooks = {
request_started = nil, --
request_finished = nil, --
},
commands = {
-- Add you custom commands here. Example:
doc = {
language_instructions = { python = "Use Google style docstrings" },
max_tokens = 1024,
},
modernize = {
user_message_template = "I have the following {{language}} code: ```{{filetype}}\n{{text_selection}}```\nModernize the above code. Use current best practices. Only return the code snippet and comments. {{language_instructions}}",
language_instructions = {
cpp = "Use modern C++ syntax. Use auto where possible. Do not import std. Use trailing return type. Use the c++11, c++14, c++17, and c++20 standards where applicable.",
},
}
},
-- Global defaults for all models
global_defaults = {
max_tokens = 4096,
temperature = 0.7,
number_of_choices = 1,
system_message_template = "You are a {{language}} coding assistant.",
user_message_template = "",
callback_type = "replace_lines",
allow_empty_text_selection = false,
extra_params = {},
max_output_tokens = nil,
},
})
setup({config}): setup pluginselect_model(): list local defined and remote available modelscancel_request(): Cancel ongoing request
Copyright (c) 2025 - Chakib Benziane [email protected]
This project is licensed under the terms of the AGPL3 license.
A copy of the license is distributed with the source code of this project.
Darby Payne (@dpayne) [email protected] the original creator.
And all contributors
