#204204th of 218 in Coding agents and AI editors
llm.nvim
Neovim plugin for ghost-text code completion backed by llm-ls
- Stars
- 1.2k
- License
- Apache-2.0
- Last commit
- Nov 2026
- Language
- Lua
1,151 on 21 Apr 20261,196 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
llm.nvim is a Lua Neovim plugin that shows Copilot-style ghost-text completions by sending requests through the llm-ls language server. It talks to the Hugging Face Inference API, Ollama, OpenAI-compatible endpoints or TGI, and trims each prompt to the configured context window using the tokenizers library. It supports fill-in-the-middle prompts for models such as StarCoder and CodeLlama.
Who it is for: Neovim users wanting self-hosted or Hugging Face-backed code completion
Strengths
- Four backends: Hugging Face Inference API, Ollama, OpenAI-compatible, TGI
- Prompts sized to context_window using real tokenizers, with a character-count fallback
- Configurable fill-in-the-middle tokens per model, with StarCoder and CodeLlama examples
- Installs via packer, lazy.nvim or vim-plug; llm-ls auto-downloaded or via Mason
Weaknesses
- Completion only; the README documents no chat, edit or agent features
- Requires the llm-ls binary, downloaded from GitHub releases on first load
- Hugging Face free Inference API tier is rate limited; PRO plan suggested
- No tagged releases; default llm-ls version pinned at 0.5.3
What it needs
- Needs Neovim, llm-ls, Hugging Face Inference API, Ollama, OpenAI-compatible server or TGI
- Models: bigcode/starcoder, bigcode/starcoder2-15b, codellama/CodeLlama-13b-hf, Ollama models, OpenAI-compatible endpoints
Also in Coding agents and AI editors
See all 218| Rank | Project | Score |
|---|---|---|
| 1 | claw-codeRust CLI coding agent harness using Anthropic and OpenAI-compatible APIs | 96 out of 100 |
| 2 | t3codeDesktop, web and mobile front end for controlling local coding agent CLIs | 95 out of 100 |
| 3 | codewhaleTerminal coding agent in Rust that works with 40+ model providers | 94 out of 100 |
| #4 | ArchonYAML workflow engine that runs AI coding agents through fixed dev steps | 93 out of 100 |
| #5 | piMinimal terminal coding agent extensible with skills and packages | 92 out of 100 |