Llama for Windows

Llama for Windows is a free, open-source Windows 11 tray app from the llama.cpp team that downloads open models from Hugging Face, picks settings that fit your PC, and gives you a local chat window plus an OpenAI-compatible API on your own machine; its new v0.13.0 adds automatic speculative decoding and a roughly 60% smaller installer.

Why it matters

For people and small teams who want ChatGPT-style help without sending documents to the cloud, it makes running open models on a Windows PC a one-click install instead of a weekend of tuning settings, and any app that speaks the OpenAI API can use the same local models.

Highlights

• Lives in the Windows system tray; Alt+Space opens a chat overlay for the loaded model from any app.
• Browse and download recommended GGUF models from the Hugging Face Hub, with fit badges (fits, might fit, won't) based on your PC's memory.
• Runs a local server your other apps can call through OpenAI-compatible endpoints, with models kept once in the shared Hugging Face cache.
• v0.13.0 detects models that ship a draft head and turns on speculative decoding automatically; the team reports up to about 2x faster output with identical results.
• The README says it sends no telemetry and prompts stay on your machine; installs shrank from 235 MB to 86 MB in v0.13.0.

Taken from the project's own README, docs or launch page. Claims about speed or accuracy are the maker's unless stated otherwise.

Worth knowing: Local models need a reasonably powerful PC; the site suggests 16 GB of memory or more for its smallest recommended model. The Microsoft Store listing was still pending certification at release.

Quick facts

Links

• Official site: https://llama.app/
• Repository: https://github.com/ggml-org/Llama-Windows
• Release notes: https://github.com/ggml-org/Llama-Windows/releases/tag/v0.13.0
• Docs: https://llama.app/docs

Related ranking

See how established tools compare in Best AI Assistants for Everyday Work. New releases aren't scored until we review them against the rubric.

Checked 2026-10-25 from the tool's own pages. Prices and star counts change, so confirm on the official site. Listing here is not a ranking or an endorsement.