gergap/vim-ollama

Vim plugin for integrating Ollama based LLM (large language models)

What it solves

It provides Copilot-like AI code completion and editing capabilities for the original Vim editor, allowing users to leverage local LLMs via Ollama to maintain privacy and avoid reliance on commercial cloud services.

How it works

The plugin uses Python scripts to communicate with the Ollama REST API (or OpenAI-compatible endpoints like LMStudio and Mistral AI) via I/O redirection. It integrates AI-generated suggestions as "ghost text" in the editor and provides a dedicated conversation buffer for chat and editing tasks. For advanced editing, it includes "OllamaEdit," which can inspect project workspaces, apply changes, and use web search/fetch tools to assist in coding.

Who it’s for

Vim users who want AI-powered coding assistance but prefer to stay within the original Vim (not NeoVim) and prioritize local model execution for privacy.

Highlights

  • Local-first AI: Uses Ollama as the primary backend for private, local code completion.
  • OllamaEdit: A tool-based workflow for AI-assisted code editing, including project workspace inspection and automatic compiler diagnostic fixes via :OllamaQuickFix.
  • Flexible Backends: Supports Ollama, OpenAI-compatible endpoints, and Mistral AI.
  • Privacy-focused: Integrates with the UNIX password manager (pass) to securely handle API keys.
  • Thinking Model Support: Displays reasoning output from thinking models in foldable sections.

Related

  • Project
    ggml-org/llama.vim

    A Vim/Neovim plugin that provides local LLM-assisted text completion and instruction-based editing via a llama.cpp server.

  • Project
    bernardo-bruning/ollama-copilot

    Ollama Copilot is a Go‑based proxy that lets you use any Ollama‑compatible (or other) LLM as a drop‑in replacement for GitHub Copilot, enabling local code completions in VS Code, Neovim, Zed, PyCharm, Emacs, etc.

  • Project
    huggingface/llm.nvim

    llm.nvim is a Neovim plugin that adds AI‑driven code completion (ghost‑text) by calling large language models via the llm‑ls language‑server. It supports multiple backends (Hugging Face Inference API, Ollama, OpenAI‑compatible servers, and Hugging  Face TGI), automatically trims prompts to fit a model’s context window, and offers configurable per‑file activation, fill‑in‑the‑middle support, and custom request parameters.

  • Project
    frankroeder/parrot.nvim

    A Neovim plugin that integrates LLM APIs for on-demand text completion, editing, and chat sessions within native buffers, emphasizing user control and privacy.

  • Project
    taketwo/llm-ollama

    A plugin for the LLM CLI tool that enables interaction with models hosted on local, remote, or cloud-based Ollama servers.