ローカルLLM
argo
Desktop AI agent platform with one-click models and offline-first RAG.
★ 829 スターCustom
ガイドを見る →Strataは、Niko1221が公開しているオープンソースのローカルLLMプロジェクトです（GitHub: Niko1221/Strata）。GitHubスターは33,338、フォークは3,565で、MITライセンスの下で公開されています。GitHubでの最終更新は2026年10月25日です。
Strata — one-click Qwen3.8-Flash-Next inference on consumer hardware: chat, code and images, locally, no account, MIT
Installers for Windows and Linux are on the GitHub releases page — the README download badge points at the latest release.
Run the installer, then launch Strata — no account, no sign-up step.
Choose a Qwen3.8-Flash-Next quantisation for your machine (ISTA-DASLab recommended first, Unsloth as an alternative), including the coder build for coding work.
A desktop installer instead of a build-from-source project — download, install, chat.
Multiple quantised builds let the same app run on consumer hardware, with a dedicated coder build for programming.
One app for conversation, coding assistance and image work, all on-device.
Free and open source, usable with your favourite coding tools, with your prompts staying local.
The README download route: the latest release on GitHub.
https://github.com/Niko1221/Strata/releases/latestThe README model guidance: ISTA-DASLab first (including the IQ1_M coder build), then Unsloth quantisations.
https://github.com/Niko1221/Strata#modelsStrata is designed to be used on the desktop and alongside the coding tools you already use, not instead of them.
https://github.com/Niko1221/StrataThe Niko1221 project on GitHub, built around the Qwen3.8-Flash-Next model family.
Yes — MIT licensed, with no account required.
Consumer hardware — the point of the quantised builds is to fit ordinary desktops and laptops; pick a smaller quantisation for weaker machines.
The README recommends ISTA-DASLab builds first, then Unsloth, with a dedicated coder build for programming work.
Download the installer from the latest GitHub release, install, launch, and pick a model build that fits your hardware.
Desktop AI agent platform with one-click models and offline-first RAG.
BitNet — Microsoft's official inference framework for 1-bit large language models, running fast on ordinary CPUs, MIT
Open bilingual (Chinese-English) 6B chat LLM with 32K-context variant and low-memory INT4 inference.
Qwen3-Coder - Alibaba's open code model series for agentic coding, up to 1M tokens of context
KoboldCpp - Run GGUF AI models locally with a KoboldAI UI. One File. Zero Install
Hundreds of models & providers. One command to find what runs on your hardware.
LibreAIカタログに掲載されたローカルLLMプロジェクトの中で最もスターの多いもの — Strataが合わなかった場合に検討したい、現実的な候補リストです。
The simplest way to run open language models on your own machine.
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
LLM inference in C/C++
A high-throughput and memory-efficient inference and serving engine for LLMs
🤯 LobeHub is your Chief Agent Operator, organizing your agents into 7×24 operations by hiring, scheduling, and reporting on your entire AI team.
カタログの中でStrataとスター数が近いローカルLLMプロジェクト — 前後の棚もそのまま見て回れます。
Run a 70B LLM on a single 4GB GPU — layer-by-layer inference with almost no memory tricks skipped
Open-source 24/7 Cowork app for OpenClaw, Hermes, Claude Code, Codex, OpenCode and 20+ more CLI Agent | Customize your assistants | Team them up｜Star if you …
Hundreds of models & providers. One command to find what runs on your hardware.
Distribute and run LLMs with a single file.
An open platform for training, serving, and evaluating large language models. Release repo for Vicuna and Chatbot Arena.
An open-source RAG-based tool for chatting with your documents.