--- title: "Local-First Directory | Software You Actually Own" url: https://www.localghost.ai/directory description: "A hand-checked list of local-first AI tools that run offline or self-host and keep your data on your own hardware, from model runtimes to personal AI." updated: 2026-11-12 site: LocalGhost.ai index: https://www.localghost.ai/llms.txt --- > THE_NETWORK # Local-First Directory Projects that run local, export clean, and don't betray the user. The crawler that will index `/.well-known/freehold.json` beacons automatically is still in development (see what's coming, below). Until it ships, this is a hand-checked starting list of local-first AI tools, grouped by where they sit in the stack. Inclusion here means a tool runs locally or self-hosts and keeps your data on your own hardware. The formal freehold verification arrives with the beacon. No ads, no rankings, no pay to play. ## 01 · Runtime the engine that runs a model on your own machine ### [Ollama](https://ollama.com) Run open models from one command, with a local API. Pulls and serves Llama, Mistral, Gemma, Qwen and more. MIT · Offline · Mac / Win / Linux ### [llama.cpp](https://github.com/ggml-org/llama.cpp) The inference engine most local tools build on. Runs quantised GGUF models on CPU or GPU. MIT · Offline · Cross-platform ### [LM Studio](https://lmstudio.ai) Desktop app to find, run and chat with local models, with an OpenAI-compatible server built in. Free / closed · Offline · Mac / Win / Linux ### [Jan](https://jan.ai) An open-source offline ChatGPT alternative. Runs models on-device with an OpenAI-compatible API. Open source · Offline · Cross-platform ### [GPT4All](https://gpt4all.io) Nomic's desktop app for running local models, including private chat over your own documents. MIT · Offline · Cross-platform ### [LocalAI](https://localai.io) A self-hosted, OpenAI-compatible API for local inference across text, audio and images. MIT · Self-host · Docker ### [vLLM](https://github.com/vllm-project/vllm) A high-throughput serving engine for running models efficiently on your own GPUs. Open source · Self-host · Linux / GPU ## 02 · Interface the window you talk to a model through ### [Open WebUI](https://openwebui.com) A self-hosted, ChatGPT-style web interface that pairs with Ollama or any OpenAI-compatible backend. Source-available · Self-host · Needs a backend ### [LibreChat](https://www.librechat.ai) A self-hostable chat interface that works across many local and remote model backends. MIT · Self-host · Needs a backend ### [AnythingLLM](https://anythingllm.com) An all-in-one app for chatting with your own documents using local models, with retrieval built in. Open source · Desktop / Docker ## 03 · Knowledge & memory where your own notes and files live, locally ### [Obsidian](https://obsidian.md) A local-first Markdown knowledge base. Your notes are plain files on your disk, with optional AI plugins. Free / closed · Offline · Cross-platform ### [Logseq](https://logseq.com) An open-source, local-first outliner that stores everything as Markdown files on your machine. Open source · Offline · Cross-platform ### [Reor](https://www.reorproject.org) A local-first AI note-taking app that runs models on-device and links related notes automatically. Open source · Offline · Desktop ## 04 · Personal AI an assistant that knows you and runs on your hardware (emerging) ### [Khoj](https://khoj.dev) A self-hostable AI that searches and chats over your own notes and documents, scoped mainly to your files. Open source · Self-host · Desktop ### [Odysseus AI](https://pewdiepie-archdaemon.github.io/odysseus/) A free, self-hosted AI workspace from PewDiePie. Chat, autonomous agents, tools, model serving, email, research and documents in one private interface that runs on your own hardware. Connect local models or your own API endpoints. Open source · Self-host · No telemetry Ours ### [LocalGhost](https://www.localghost.ai) A fleet of small Go daemons for notes, memory, voice, vision and adversarial review, running on hardware you own with no cloud dependency. Early-stage and in active development, not yet a finished product. MIT · Self-host · In development ## Specialised local-first tools for one domain, beside the assistant stack ### [Rotki](https://rotki.com) A local-first, privacy-preserving portfolio and accounting tracker. Your financial data stays on your machine. Open source · Offline · Desktop ### [whisper.cpp](https://github.com/ggml-org/whisper.cpp) Fully offline speech-to-text on your own hardware. Transcribe audio with no upload. MIT · Offline · Cross-platform ### [ComfyUI](https://www.comfy.org) A node-based interface for running image-generation models locally, with full control of the pipeline. Open source · Offline · GPU ## > WHAT'S COMING - Automated discovery of projects hosting `/.well-known/freehold.json` - Submit your project URL and we'll validate your beacon file - A searchable index by capability (offline-first, self-hosted, export formats) - Verification that your project works offline and the export works - RSS feed to follow new additions to the index - Full directory downloadable as JSON, take the data, run it yourself ## > WHAT STILL NEEDS TO BE BUILT See something missing from the ecosystem? We've mapped the gaps, the critical infrastructure, the tools that don't exist yet, and where help is needed most. [VIEW THE ROADMAP →](https://www.localghost.ai/build) [← THE MANIFESTO](https://www.localghost.ai/manifesto) [WHY LOCAL AI →](https://www.localghost.ai/why-local-ai) --- Cite this page as https://www.localghost.ai/directory. Facts about LocalGhost are kept current at https://www.localghost.ai/about, and the index of the whole site for language models is https://www.localghost.ai/llms.txt.