Meilleurs outils IA auto-hébergés (2026)
Exécutez l'IA en local en toute confidentialité. LLM open source, interfaces de chat, bases de connaissances et outils de dev à auto-héberger sur votre infrastructure.
Meetily — Privacy-First AI Meeting Assistant with Local Transcription
An open-source, self-hosted AI meeting assistant that provides real-time transcription, speaker diarization, and local summarization using Whisper and Ollama, with no cloud dependency.
Self-Hosted AI Starter Kit — Local AI with n8n
Docker Compose template by n8n that bootstraps a complete local AI environment with n8n workflow automation, Ollama LLMs, Qdrant vector database, and PostgreSQL. 14,500+ stars.
Continue — Open-Source AI Code Assistant
Open-source AI code assistant for VS Code and JetBrains. Tab autocomplete, chat, inline editing with any model — OpenAI, Anthropic, Ollama, or self-hosted.
Colanode — Open-Source Local-First Slack and Notion Alternative
A self-hosted collaboration platform combining real-time team chat, a rich document editor, and a knowledge base in a single local-first application that keeps all data on your own infrastructure.
ArchiveBox — Self-Hosted Web Archiving Platform
ArchiveBox is an open-source self-hosted web archiver that saves URLs as local HTML, PDF, screenshots, WARC, and more. Feed it bookmarks, browser history, or RSS feeds and it preserves everything for offline access.
Open CoDesign — Open-Source AI Design Tool with Multi-Model Support
A local-first, open-source alternative to commercial AI design tools that generates prototypes, slides, and PDFs from prompts using Claude, GPT, Gemini, or local models via Ollama.
Ollama Model Library — Best AI Models for Local Use
Curated guide to the best models available on Ollama for coding, chat, and reasoning. Compare Llama, Mistral, Gemma, Phi, and Qwen models for local AI development.
Audiobookshelf — Self-Hosted Audiobook & Podcast Server
Audiobookshelf is an open-source audiobook and podcast server with progress sync, chapter navigation, mobile apps, and multi-user support — a self-hosted Audible alternative.
Actual Budget — Local-First Personal Finance App
Actual is an open-source personal finance app with envelope budgeting, bank sync, multi-device sync, and local-first architecture — a YNAB alternative.
Immich — High-Performance Self-Hosted Photo & Video Management
Immich is an open-source Google Photos alternative with auto-backup, AI-powered search, face recognition, and mobile apps — self-hosted for complete privacy.
Verba — The Golden RAGtriever by Weaviate
Verba is an open-source RAG (Retrieval-Augmented Generation) chatbot from the Weaviate team. Drop in PDFs, web pages, or notes; pick a model (OpenAI, Ollama, Anthropic); and get a polished chat UI with semantic search built in.
Stirling PDF — Self-Hosted PDF Editor & Toolkit
Stirling PDF is the #1 open-source PDF tool on GitHub. Merge, split, convert, compress, OCR, sign, and edit PDFs — all self-hosted with no data leaving your server.
Open WebUI — Self-Hosted AI Chat Interface
User-friendly, self-hosted AI chat interface. Supports Ollama, OpenAI, Anthropic, and any OpenAI-compatible API. RAG, web search, voice, image gen, and plugins. 129K+ stars.
Firefly III — Self-Hosted Personal Finance Manager
Firefly III is an open-source personal finance manager for tracking expenses, budgets, and bank accounts. Self-hosted with full privacy, multi-currency, and powerful reporting.
HuggingFace Chat UI — Open-Source AI Chat Interface
Chat UI is Hugging Face's open-source web interface for conversational AI, powering HuggingChat and supporting any text-generation model via TGI, Ollama, or OpenAI-compatible APIs with features like web search, tool use, and multimodal input.
Ollama — Run LLMs Locally
Run large language models locally on your machine. Supports Llama 3, Mistral, Gemma, Phi, and dozens more. One-command install, OpenAI-compatible API.
/loop — Local Recurring Task Scheduler (Boris-Style)
Open-source slash command for recurring local Claude Code tasks with a 3-day safety cap. Inspired by Boris Cherny's /loop scheduler.
Vikunja — Self-Hosted Open Source To-Do and Project Management
Vikunja is an open-source task management application written in Go. It offers lists, kanban boards, Gantt charts, and CalDAV sync — a self-hosted alternative to Todoist, Trello, and Asana.
LocalAI — Run Any AI Model Locally, No GPU
LocalAI is an open-source AI engine running LLMs, vision, voice, and image models locally. 44.6K+ GitHub stars. OpenAI/Anthropic-compatible API, 35+ backends, MCP, agents. MIT licensed.
LobeChat — Open-Source Multi-Model Chat UI
Beautiful open-source chat UI supporting Claude, GPT-4, Gemini, Ollama, and 50+ providers. Plugin system, knowledge base, TTS, image generation, and self-hostable. 55,000+ GitHub stars.
Halo — Modern Self-Hosted Publishing Platform
Halo is an open-source content management and blogging platform built with Java and Spring Boot. It provides a polished editing experience, a plugin system, and theme marketplace for self-hosted publishing.
Local Deep Research — Privacy-First AI Research Agent
A self-hosted deep research agent that achieves near-perfect accuracy on benchmarks using local or cloud LLMs, with support for 10+ search engines and fully encrypted processing.
Tolgee — Developer-Friendly Localization Platform
An open-source localization platform that lets developers and translators manage translations through a web UI, in-context editing, and native SDK integrations for React, Vue, Angular, and more.
Cherry Studio Custom Models — BYOK Any LLM Provider
Cherry Studio Custom Models adds any OpenAI-compatible endpoint — proxy, local, or third-party. Mix Claude, GPT, Gemini, DeepSeek, Ollama side-by-side.
LibrePhotos — Self-Hosted AI-Powered Photo Management
A self-hosted open-source photo management service with automatic face recognition, object detection, and geolocation tagging powered by machine learning.
Home Assistant — Open-Source Home Automation That Puts Local Control First
Home Assistant is the most popular open-source platform for smart home automation. It integrates 3,000+ devices and services, runs entirely on local hardware (Raspberry Pi to NUC), and keeps your data off the cloud by default.
Refact — Local-First AI Coding Assistant
Refact is an open-source, local-first AI coding assistant: install the IDE plugin, run local refact-lsp, and connect a model provider.
SiYuan — Privacy-First Self-Hosted Knowledge Management
SiYuan is a local-first, self-hosted personal knowledge management system with block-level references, end-to-end encryption, and Markdown support.
frp — Fast Reverse Proxy to Expose Local Servers Behind NATs and Firewalls
frp is a high-performance reverse proxy written in Go. Expose a local HTTP/TCP/UDP service to the public internet through a relay server — the self-hosted alternative to ngrok and Cloudflare Tunnel.
Stable Diffusion Web UI by AUTOMATIC1111 — The Definitive Local AI Image Generator
AUTOMATIC1111's Stable Diffusion Web UI is the most popular interface for running Stable Diffusion locally. It supports text-to-image, image-to-image, inpainting, ControlNet, LoRA, embeddings, extensions, and every model variant — all in a self-hosted browser UI.
La stack IA auto-hébergée
The Self-Hosted AI Stack
Self-hosted AI has matured from a hobbyist pursuit to an enterprise requirement. Privacy regulations, data sovereignty laws, and the desire for predictable costs drive organizations to run AI on their own infrastructure. Local LLM Inference — Ollama, Jan, and GPT4All make running models like Llama, Mistral, and Qwen as simple as installing an app. Support for GPU acceleration, quantization, and model management.
Chat Interfaces — Open WebUI, LibreChat, LobeChat, and AnythingLLM provide ChatGPT-like interfaces for your self-hosted models. Features include conversation history, file upload, RAG integration, and multi-model switching. Knowledge Bases — Onyx, Quivr, and PrivateGPT let you build private RAG systems over your documents — no data leaves your servers.
Development Tools — Tabby (self-hosted Copilot), SearXNG (private search), and Puter (cloud desktop) provide developer infrastructure without external dependencies. TokRepo hosts deployment configs and Docker Compose files for the entire self-hosted AI stack.
Self-hosting AI isn't about avoiding costs — it's about owning your intelligence infrastructure.
Questions fréquentes
Quel matériel faut-il pour auto-héberger de l'IA ?+
Cela dépend de la taille du modèle. Pour des modèles 7B paramètres (suffisants pour la plupart des tâches) : 16 Go de RAM + tout GPU moderne avec 8 Go de VRAM. Pour des modèles 70B (classe GPT-4) : 64 Go de RAM + GPU avec 48 Go de VRAM (A6000 ou double 3090). Pour de l'inférence CPU uniquement : Ollama avec des modèles quantisés tourne sur n'importe quel laptop moderne, juste plus lentement. Les Mac Apple Silicon avec 32 Go+ de mémoire unifiée sont excellents pour l'IA locale.
L'IA auto-hébergée vaut-elle les API cloud ?+
Pour de nombreuses tâches, oui. Les modèles open source comme Llama 3.1 70B et Qwen 2.5 72B rivalisent avec GPT-4 en code, analyse et raisonnement général. Ils restent en retrait sur le raisonnement multi-étapes le plus complexe et les tâches créatives où Claude Opus ou GPT-4o restent en tête. L'écart se réduit chaque trimestre. Pour la plupart des applications business, les modèles auto-hébergés sont "suffisants" avec une bien meilleure confidentialité et un coût bien inférieur.
Quelle est la façon la plus simple de débuter avec l'IA auto-hébergée ?+
Installez Ollama (une seule commande sur Mac/Linux/Windows), pullez un modèle ("ollama pull llama3.1"), puis installez Open WebUI pour une interface façon ChatGPT. Temps de setup total : moins de 10 minutes. TokRepo héberge des configs Docker Compose qui regroupent Ollama + Open WebUI + pipeline RAG dans un seul déploiement.