#157157th of 197 in Coding agents and AI editors

llama.vscode

VS Code extension for local FIM completion, chat and agentic coding on llama.cpp

Stars
1.6k
License
MIT
Last commit
Nov 2026
Last release
Nov 2026
Language
TypeScript
Stars, last 29 weeks+312
21 Apr 2026: 1,241 stars28 Apr 2026: 1,254 stars5 May 2026: 1,275 stars12 May 2026: 1,292 stars19 May 2026: 1,311 stars26 May 2026: 1,326 stars2 Jun 2026: 1,337 stars9 Jun 2026: 1,353 stars16 Jun 2026: 1,368 stars23 Jun 2026: 1,379 stars30 Jun 2026: 1,396 stars7 Jul 2026: 1,401 stars14 Jul 2026: 1,417 stars21 Jul 2026: 1,432 stars28 Jul 2026: 1,444 stars4 Aug 2026: 1,449 stars11 Aug 2026: 1,462 stars18 Aug 2026: 1,466 stars25 Aug 2026: 1,475 stars1 Sept 2026: 1,482 stars8 Sept 2026: 1,490 stars15 Sept 2026: 1,498 stars22 Sept 2026: 1,506 stars29 Sept 2026: 1,514 stars6 Oct 2026: 1,523 stars13 Oct 2026: 1,531 stars20 Oct 2026: 1,543 stars27 Oct 2026: 1,548 stars3 Nov 2026: 1,550 stars10 Nov 2026: 1,553 stars

1,241 on 21 Apr 20261,553 on 10 Nov 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

llama.vscode is a VS Code extension that sends fill-in-the-middle completion requests to a local llama.cpp server, with a ring context built from open and edited files. It also includes a chat view and Llama Agent, which works with local models or external ones such as OpenRouter and can use MCP servers installed in VS Code. Models are grouped into envs that can be selected, imported and exported together.

Who it is for: VS Code users wanting local code completion and an agent on llama.cpp

Strengths

  • Completion runs locally on llama.cpp, with documented settings for under 8GB up to 64GB+ VRAM
  • Ring context from open files and yanked text, with cache reuse for large contexts
  • Envs switch whole model sets (completion, chat, embeddings, tools) in one step
  • Agent can use MCP servers installed in VS Code and external providers like OpenRouter

Weaknesses

  • Completion requires FIM-compatible models, which limits model choice
  • CPU-only setups are documented as significantly lower quality
  • Tied to VS Code; Vim/Neovim needs the separate llama.vim plugin
  • Agent ships 9 internal tools; custom tools must be written in JavaScript

What it needs

  • GPU optional
  • Needs VS Code, llama.cpp
  • Models: GGUF, FIM models (Qwen2.5-Coder), gpt-oss 20B, OpenRouter, Hugging Face downloads
  • port 8012

Quick answers

Is llama.vscode open source?

Yes. Its code is on GitHub under the MIT license.

Also in Coding agents and AI editors

See all 197
Also in Coding agents and AI editors
RankProjectScore
1claw-codeRust CLI coding agent harness using Anthropic and OpenAI-compatible APIs194.8k stars+739 (+0.4%) in 30 daysMIT97 out of 100
2t3codeDesktop, web and mobile front end for controlling local coding agent CLIs Live demo ↗ (opens in a new tab)56.5k stars+29,061 (+106%) in 30 daysMIT94 out of 100
3codewhaleTerminal coding agent in Rust that works with 40+ model providers41.2k stars+334 (+0.8%) in 30 daysMIT94 out of 100
#4piMinimal terminal coding agent extensible with skills and packages121.4k stars+7,392 (+6.5%) in 30 daysMIT92 out of 100
#5oh-my-openagentTerminal coding agent with multi-model orchestration and git-backed memory70.8k stars+1,253 (+1.8%) in 30 dayscustom license92 out of 100