AlexsJones / llmfit

📊 New: benchmark & share — real numbers from your machine, better estimates for everyone. Download a model, serve it, and measure real tok/s on your hardware — then contribute the results back to the project as a PR, straight from the TUI. No gh CLI, no third-party account. Every run is saved locally first, your own measurements replace estimates in the fit table, and each merged submission ships in the next release: anyone on identical hardware gets measured ✓ numbers before they ever run a benchmark. Follow the step-by-step benchmarking guide → 📊 新功能：基准测试与分享 — 获取来自你机器的真实数据，为所有人提供更准确的估算。下载模型、运行服务，并测量你硬件上的真实 tok/s（每秒 token 数），然后直接通过 TUI（终端用户界面）将结果作为 PR 贡献回项目。无需 gh CLI，无需第三方账户。每次运行都会先保存在本地，你的测量数据将替换适配表中的估算值，且每次合并的提交都会包含在下个版本中：任何使用相同硬件的用户在运行基准测试前，就能直接看到已验证的 ✓ 数据。请参考分步基准测试指南 →

Previously: llmfit 1.0 — the release where the numbers became verifiable → Hundreds of models & providers. One command to find what runs on your hardware. A terminal tool that right-sizes LLM models to your system’s RAM, CPU, and GPU. Detects your hardware, scores each model across quality, speed, fit, and context dimensions, and tells you which ones will actually run well on your machine. Ships with an interactive TUI (default) and a classic CLI mode. Supports multi-GPU setups, MoE architectures, dynamic quantization selection, speed estimation, and local runtime providers (Ollama, llama.cpp, MLX, Docker Model Runner, LM Studio). 此前：llmfit 1.0 — 在该版本中，数据变得可验证 → 数百个模型与提供商。只需一条命令即可找到适合你硬件运行的模型。这是一个能根据系统 RAM、CPU 和 GPU 调整 LLM 模型大小的终端工具。它能检测你的硬件，从质量、速度、适配度和上下文维度对每个模型进行评分，并告诉你哪些模型在你的机器上能真正运行良好。自带交互式 TUI（默认）和经典 CLI 模式。支持多 GPU 设置、MoE 架构、动态量化选择、速度估算以及本地运行时提供商（Ollama、llama.cpp、MLX、Docker Model Runner、LM Studio）。

Sister projects: sympozium — managing agents in Kubernetes. llmserve — a simple TUI for serving local LLM models. Pick a model, pick a backend, serve it. llama-panel — a native macOS app for managing local llama-server instances. 姊妹项目：sympozium — 在 Kubernetes 中管理智能体。llmserve — 一个用于托管本地 LLM 模型的简单 TUI。选择模型、选择后端，即可运行。llama-panel — 一个用于管理本地 llama-server 实例的 macOS 原生应用。

Documentation

文档

Get started: Install · Usage · How it works 入门：安装 · 使用方法 · 工作原理

Guides: TUI guide · Benchmarking step-by-step · CLI & automation · Runtime providers · OpenClaw integration 指南：TUI 指南 · 分步基准测试 · CLI 与自动化 · 运行时提供商 · OpenClaw 集成

Reference: How it works (full) · Platform & GPU support · Custom models · Development 参考：工作原理（完整版） · 平台与 GPU 支持 · 自定义模型 · 开发

Project: Contributing · Alternatives · Code signing · License 项目：贡献 · 替代方案 · 代码签名 · 许可证

Install

安装

Windows scoop install llmfit If Scoop is not installed, follow the Scoop installation guide. 如果未安装 Scoop，请参考 Scoop 安装指南。

macOS / Linux Homebrew Prebuilt binary (recommended, works on all macOS/Linux versions): brew install AlexsJones/llmfit/llmfit Or from the homebrew-core formula, which builds from source on macOS versions without a bottle: brew install llmfit 预编译二进制文件（推荐，适用于所有 macOS/Linux 版本）： brew install AlexsJones/llmfit/llmfit 或者使用 homebrew-core 公式（在没有 bottle 的 macOS 版本上会从源码构建）： brew install llmfit

MacPorts port install llmfit

Quick install curl -fsSL https://llmfit.axjns.dev/install.sh | sh Downloads the latest release binary from GitHub and installs it to /usr/local/bin (or ~/.local/bin if no sudo). Install to ~/.local/bin without sudo: curl -fsSL https://llmfit.axjns.dev/install.sh | sh -s -- --local 快速安装 curl -fsSL https://llmfit.axjns.dev/install.sh | sh 从 GitHub 下载最新发布版二进制文件并安装到 /usr/local/bin（若无 sudo 则安装到 ~/.local/bin）。 无需 sudo 安装到 ~/.local/bin： curl -fsSL https://llmfit.axjns.dev/install.sh | sh -s -- --local

uv / pip To install or update llmfit: uv tool install -U llmfit To run without installing: uvx llmfit You can also install llmfit as a Python package in the normal way with tools such as pip or uv. uv / pip 安装或更新 llmfit： uv tool install -U llmfit 无需安装直接运行： uvx llmfit 你也可以通过 pip 或 uv 等工具以常规方式将 llmfit 作为 Python 包安装。

Docker / Podman docker run ghcr.io/alexsjones/llmfit This prints JSON from llmfit recommend command. The JSON could be further queried with jq. podman run ghcr.io/alexsjones/llmfit recommend --use-case coding | jq '.models[].name' To launch the interactive TUI instead, pass the global —tui flag: docker run --rm -it ghcr.io/alexsjones/llmfit --tui Docker / Podman docker run ghcr.io/alexsjones/llmfit 这将打印来自 llmfit recommend 命令的 JSON。该 JSON 可以通过 jq 进一步查询。 podman run ghcr.io/alexsjones/llmfit recommend --use-case coding | jq '.models[].name' 若要启动交互式 TUI，请传入全局 —tui 标志： docker run --rm -it ghcr.io/alexsjones/llmfit --tui

From source git clone https://github.com/AlexsJones/llmfit.git cd llmfit cargo build --release # binary is at target/release/llmfit 从源码构建 git clone https://github.com/AlexsJones/llmfit.git cd llmfit cargo build --release # 二进制文件位于 target/release/llmfit

Usage

使用方法

llmfit # interactive TUI: your hardware, every model, ranked llmfit # 交互式 TUI：显示你的硬件，以及所有模型的排名

The TUI shows your detected specs at the top and every model scored for fit, speed, quality, and context. See the TUI guide for navigation, planning, simulation, downloads, the community leaderboard, and benchmarking. TUI 在顶部显示检测到的规格，并对每个模型的适配度、速度、质量和上下文进行评分。请参阅 TUI 指南以了解导航、规划、模拟、下载、社区排行榜和基准测试。

For scripts, agents, and classic terminal output: 针对脚本、智能体和经典终端输出：

llmfit fit # table of all models ranked by fit llmfit fit # 按适配度排名的所有模型表格

llmfit recommend --json # top picks as JSON (agent/script consumption) llmfit recommend --json # 以 JSON 格式输出推荐模型（供智能体/脚本调用）

llmfit info "<model>" # one model: fit analysis, estimate basis, verify commands llmfit info "<model>" # 单个模型：适配度分析、估算依据、验证命令

llmfit bench # measure real tok/s/TTFT against your running provider llmfit bench # 测量当前运行提供商的真实 tok/s/TTFT

llmfit doctor # hardware detection report for bug reports llmfit doctor # 用于错误报告的硬件检测报告

Full reference: CLI & automation. 完整参考：CLI 与自动化。

How it works

工作原理

llmfit detects your hardware (RAM, CPU, GPU/VRAM, backend), then scores every model in its catalog across four dimensions: memory fit, estimated speed, quality, and context. Speed estimates come from a memory-bandwidth model grounded in runtime sampling and real community measurements — and every estimate ships its inputs, so llmfit info shows exactly what a number assumes and how to verify it on your machine. Full detail, including the estimation formulas and the model database: How llmfit works. llmfit 检测你的硬件（RAM、CPU、GPU/VRAM、后端），然后从四个维度对目录中的每个模型进行评分：内存适配度、预估速度、质量和上下文。速度估算基于运行时采样和真实社区测量得出的内存带宽模型 — 且每个估算都附带其输入数据，因此 llmfit info 会准确显示该数值的假设前提以及如何在你的机器上进行验证。完整细节（包括估算公式和模型数据库）：How llmfit works。

Contributing

贡献

Contributions are welcome, especially new models. Before submitting a PR Please run cargo fmt before pushing your changes. Most CI check failures are caused by unformatted code: cargo fmt 欢迎贡献，特别是新模型。在提交 PR 之前，请在推送更改前运行 cargo fmt。大多数 CI 检查失败都是由未格式化的代码引起的：cargo fmt

Guides for adding models — locally (no rebuild) or to the built-in catalog: Custom models. 添加模型的指南 — 本地添加（无需重新构建）或添加到内置目录：Custom models。

Alternatives

替代方案

If you’re looking for a different approach, check out llm-checker — a Node.js CLI tool with Ollama integration that can pull and benchmark models directly. It takes a more hands-on approach by actually running models on your hardware via Ollama, rather than estimating from specs. Good if you already have Ollama installed and want to test real-world performance. Note that it doesn’t support MoE (Mixture-of-Experts) architectures — all models are treated as dense, so memory estimates for models like Mixtral or DeepSeek-V3 will reflect total parameter count rather than the smaller active subset. 如果你在寻找不同的方法，可以看看 llm-checker —— 一个带有 Ollama 集成的 Node.js CLI 工具，可以直接拉取并对模型进行基准测试。它采取更直接的方法，通过 Ollama 在你的硬件上实际运行模型，而不是根据规格进行估算。如果你已经安装了 Ollama 并想测试真实性能，这很适用。注意它不支持 MoE（混合专家）架构 —— 所有模型都被视为稠密模型，因此像 Mixtral 或 DeepSeek-V3 这类模型的内存估算将反映总参数量，而不是较小的活跃子集。

Code signing

代码签名

llmfit’s Windows release binaries are digitally signed (Authenticode) via SignPath.io, with a free code signing certificate provided by the SignPath Foundation. Signing happens automatically in the release pipeline: only artifacts built by GitHub Actions from this repository are submitted for signing, and signing requests are approved by the project maintainer (@AlexsJones). Code signing policy: see the SignPath Foundation code signing policy and terms. llmfit 的 Windows 发布版二进制文件通过 SignPath.io 进行数字签名（Authenticode），并使用由 SignPath Foundation 提供的免费代码签名证书。签名在发布流水线中自动完成：只有从本仓库通过 GitHub Actions 构建的制品才会被提交签名，且签名请求由项目维护者 (@AlexsJones) 批准。代码签名政策：请参阅 SignPath Foundation 代码签名政策和条款。

Privacy: this program will not transfer any information to other networked systems unless specifically requested by the user or the person installing or operating it. llmfit only contacts external services when you explicitly use the corresponding feature (e.g. model downloads, runtime provider queries, or the community leaderboard). 隐私：除非用户或安装/操作人员明确要求，否则本程序不会将任何信息传输到其他网络系统。llmfit 仅在你明确使用相应功能（例如模型下载、运行时提供商查询或社区排行榜）时才会联系外部服务。