Skip to content
Get packs

llmfit

A terminal tool that shows which open-source LLMs your computer can run well.

GitHub Download v1.1.18
Stars
38.6k
Forks
2.5k
Licence
MIT
Language
Rust
Last update
4 days ago
Type
CLI

What it is

llmfit checks your CPU, RAM, GPU and VRAM, then recommends open-source large language models that should run comfortably on that hardware. It scores models on fit, speed, quality and context length, and can show results in an interactive terminal interface, a plain CLI, or a web dashboard. People use it to avoid downloading models their machine cannot handle.

  • Run llmfit recommend to see a ranked list of models for your hardware
  • Use llmfit recommend --json with jq to filter model names for a use case such as coding
  • Simulate different hardware with the TUI to plan a deployment
  • Run llmfit serve to expose a local HTTP API with /api/v1/system and /api/v1/models endpoints
  • Benchmark a downloaded model and save results locally

Great for

  • People deciding which local LLM to download for their laptop or desktop
  • Developers comparing quantized models (GGUF, AWQ, GPTQ, EXL2) for a specific machine
  • Teams sizing hardware or SSD space for a set of local models
  • Users of local runtimes like Ollama, llama.cpp, MLX, LM Studio or Docker Model Runner

Not ideal for

  • People who only use hosted APIs and never run models locally, since the tool is about local hardware fit
  • Users who need exact speed numbers without benchmarking, because the README says estimates are projections and measured results are optional
  • Anyone needing a signed Windows binary by default, since the README says some releases may ship unsigned Windows artifacts
Step-by-step guide: how to use llmfitInstall, first run, examples, tips and FAQs

How to get it

  1. 1

    Install with a package manager

    Install on Windows with WinGet

    winget install --exact --id AlexsJones.llmfit

    Install with Homebrew

    brew install AlexsJones/llmfit/llmfit

    Install with uv tool

    uv tool install -U llmfit
  2. 2

    Run it

    • Install llmfit using one of the package manager commands listed in the README
    • Run llmfit in a terminal to open the interactive TUI
    • Press / to filter models by name, family or quantization
    • Run llmfit recommend for plain-text or llmfit recommend --json output
  3. 3

    Get the most out of it

    • Use llmfit recommend --json with jq to script model selection
    • Read the TUI guide (docs/tui.md) for planning and simulation features
    • Follow the benchmarking guide (docs/benchmarking.md) to replace estimates with measured tok/s on your hardware
    • Use llmfit serve to connect the REST API to dashboards or automation
    • Check docs/providers.md to see which local runtimes are supported

Downloads

Source code (.zip)main branch · straight from GitHubRelease v1.1.188 files incl. llmfit-v1.1.18-x86_64-unknown-linux-musl.tar.gz

Strengths

  • Detects hardware automatically, so you do not have to look up your own specs
  • Offers a TUI, a CLI with JSON output, a web dashboard and an HTTP API
  • Available through several package managers including Homebrew, WinGet, Scoop, uv/pip and Docker

Trade-offs

  • Recommendations depend on estimates for memory and speed unless you measure your own hardware
  • The README notes that Windows release binaries are not always signed
  • Platform support is listed as macOS, Linux (x86_64 and ARM64) and Windows (x86_64), so other systems may not be covered

How it compares

ProjectStarsTypeLevelUpdatedBest for
llmfit (this)38.6kCLIbeginner4 days agoPeople deciding which local LLM to download for their laptop or desktop
mlx29kLibraryintermediate4 days agoResearchers experimenting with machine learning models on a Mac with Apple silicon
anything-llm67.8kAppbeginner6 days agoPeople who want a private ChatGPT-style assistant over their own PDFs and text files
voicebox59.2kAppbeginner30 days agoPeople who want text-to-speech or voice cloning without sending audio to cloud services
patent-disclosure-skill12.7kAI model / agentintermediate4 days agoEngineers who have technical work and want to draft a Chinese patent disclosure document
llama-gpt10.9kAppintermediate2.5 years agoPeople who want a private chatbot and don't want to send prompts to a cloud provider

Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.