How to set up vim-llm-agent with local Ollama models?

Install Ollama locally, then set g:llm_agent_provider to 'ollama' and specify a model like 'llama3.2' in your .vimrc. The plugin defaults to localhost:11434, but you can customize the base URL if needed.

vim-llm-agent vs Copilot for Vim: which is better for code assistance?

vim-llm-agent offers multi-provider support and autonomous tools with plan approval, ideal for flexible, context-aware tasks. Copilot provides tighter GitHub integration and completions but is limited to OpenAI; choose based on whether you need control or seamless suggestions.

Can vim-llm-agent edit files automatically without confirmation?

Yes, but only if you disable plan approval by setting g:llm_agent_require_plan_approval to 0. By default, it requires user approval for each step to ensure safety and transparency in file modifications.

How does the conversation summary compaction work?

It automatically compresses older conversation into a summary when history exceeds 50KB (configurable), keeping recent 20KB uncompressed. This manages token limits while preserving context, with manual triggers via :GptGenerateSummary.

Is vim-llm-agent free to use with all providers?

The plugin is open-source, but most providers like OpenAI and Anthropic charge API fees. Ollama allows free local usage, and OpenRouter has varied pricing, so costs depend on your chosen model and usage.

How to troubleshoot the 'ModuleNotFoundError: No module named requests' error?

Install requests for the specific Python version Vim uses, often requiring commands like 'python3.13 -m pip install requests' with --break-system-packages or --user flags, as detailed in the troubleshooting section for compatibility issues.

Open-Awesome

vim-chatgpt

UnlicensePython

A Vim plugin that integrates multiple AI providers (OpenAI, Anthropic, Gemini, Ollama, OpenRouter) for code assistance with autonomous tool execution.

GitHub

272 stars35 forks0 contributors

What is vim-chatgpt?

vim-llm-agent is a Vim plugin that integrates multiple large language model providers directly into the Vim editor for AI-powered code assistance. It enables developers to get code explanations, reviews, refactoring suggestions, and documentation while maintaining conversation history and project context. Unlike simple chat interfaces, it features autonomous tool execution for file operations and git commands with user approval workflows.

Target Audience

Vim users and developers who want AI assistance directly in their editor workflow, particularly those working on codebases who need explanations, reviews, refactoring help, or automated file operations.

Value Proposition

Developers choose vim-llm-agent for its multi-provider flexibility, autonomous tool execution with plan approval, and deep Vim integration that maintains project context and conversation history across sessions.

Overview

Vim Plugin For ChatGPT

Use Cases

Best For

Getting AI-powered code explanations and reviews directly in Vim
Refactoring code with autonomous file editing and plan approval
Generating documentation and tests for selected code snippets
Exploring project structure using AI-driven file navigation
Creating commit messages based on buffer changes
Maintaining context-aware conversations across coding sessions

Not Ideal For

Developers using GUI-based editors like VS Code or Sublime Text who want integrated AI tools
Teams with high security requirements that prohibit autonomous file editing by AI agents
Projects where minimal Vim configuration is desired, as it requires Python support and external dependencies
Users who need simple, one-off AI queries without the overhead of API key setup and plugin installation

Pros & Cons

Pros

Multi-Provider Flexibility

Supports OpenAI, Anthropic, Gemini, Ollama, and OpenRouter, allowing users to switch providers based on cost, performance, or privacy needs, as detailed in the 'Supported Providers' section.

Autonomous Tool Execution

Enables AI to perform file operations, git commands, and project exploration with plan approval workflows, making complex refactoring tasks efficient and transparent, as described in the 'AI Tools & Function Calling' section.

Context-Aware Conversations

Maintains conversation history and project context across sessions with automatic summary compaction, ensuring continuity and bounded token usage, as explained in the 'Conversation Summary & Preferences' section.

Comprehensive Code Commands

Provides dedicated commands for explaining, reviewing, refactoring, documenting, testing, and fixing code, directly integrated into Vim's visual selection workflow for seamless assistance.

Cons

Python Dependency Issues

Requires Vim with Python3 support and the requests library, leading to installation headaches like ModuleNotFoundError, as highlighted in the troubleshooting section with complex workarounds for different Python versions.

Limited Tool Support

Tool calling is only supported by OpenAI and Anthropic models, excluding Gemini and Ollama from autonomous file operations, which reduces flexibility for users preferring other providers.

Complex Configuration

Setting up involves numerous Vimscript variables for providers, models, and workflow options, which can be overwhelming for users unfamiliar with Vim configuration or AI API integrations.

Frequently Asked Questions

Related Projects

PandasAI

Chat with your database or your datalake (SQL, CSV, parquet). PandasAI makes data analysis conversational using LLMs and RAG.

Stars23,660

Forks2,336

Last commit8 months ago

chatgpt-vscode

A VSCode extension that allows you to use ChatGPT

Stars4,930

Forks358

Last commit2 years ago

AICommand

ChatGPT integration with Unity Editor

Stars4,109

Forks424