Self-Hosted AI Tools You Can Run in Docker

Unstract is a self-hosted, no-code platform for teams that need structured data from documents whose layouts vary. It uses LLMs to extract JSON from PDFs, images, scans and office files without building a template for each vendor or form. Its use cases include finance, insurance, healthcare and KYC document processing.

next-wiki is a self-hosted wiki for people and teams who want AI-assisted knowledge management with a record of where information came from. It keeps original sources separate from AI-written material and reviewed pages, so an assistant's synthesis remains traceable to its evidence. It's open source under Apache 2.0 and runs as a browser app on your own Docker host with PostgreSQL.

Chatwoot is a self-hosted customer support platform for teams that want conversation records and customer data on their own server. Its optional paid Captain AI connects to OpenAI by default, or to a self-hosted OpenAI-compatible model endpoint. Captain sends data to the model service you choose, so hosting the support desk and hosting its AI are separate decisions.

Ollama Helm Chart packages a self-hosted LLM service for developers and infrastructure teams that already use Kubernetes. It brings Ollama into a cluster so applications can access models running on the cluster's hardware.

KTransformers is a research framework for developers running or fine-tuning large language models on their own hardware. It focuses on mixture-of-experts (MoE) models, using CPU resources alongside GPUs to work within limited GPU memory.

HelixDB is a database for developers building AI agent memory and retrieval-augmented generation (RAG). It keeps entities, relationships, embeddings, and text together so applications can search both content and its connections.

pgvector adds vector similarity search to Postgres for developers building applications that need to find related items. It keeps embeddings alongside application records, so vector searches can use the same database as the rest of your data.

haiku.rag lets you search your own documents and ask questions with answers you can check against the original pages. It's for people building a private document library or giving an AI assistant access to their files.

Mirrai is an AI companion for people who want to revisit unfinished conversations through a persona based on someone's chat history. It simulates their communication style, with a farewell process that gives the experience an optional endpoint.

InvokeAI is a free, open-source image generation application for artists who want to create and edit artwork on their own hardware. Its locally hosted browser interface combines direct canvas work with repeatable generation workflows.

Dograh is an open source voice AI platform for teams building agents that answer or place phone calls. It's a self-hosted alternative to Vapi and Retell, with control over conversation flows and a choice of AI providers.

Xinference serves AI models on your own computer or servers for developers and researchers building applications around models they control. It brings chat, document search and media generation into one platform.

Kestra coordinates AI, data and infrastructure workflows for teams that need to connect tasks and track their execution. It can run existing scripts and containers, so workflows can span tools written in different languages.

SwanLab is an open-source AI experiment tracking platform for researchers and model training teams. It brings training results into one place so you can spot problems and compare runs before choosing your next experiment.

Voice Changer (w-okada), also called VC Client, converts speech as you talk using AI voice conversion models. It suits people who want live voice changes, including gaming commentators running other demanding applications.

NeMo Guardrails is a Python toolkit for developers building AI assistants that need rules for what they say and which actions they take. It adds checks around conversations, document-based answers and tool use.

OwnPilot is a self-hosted AI assistant for people who want personal records and agents that can carry out tasks in one web interface. It supports local model backends as well as cloud providers.

whisper.cpp runs OpenAI's Whisper speech recognition models on your own hardware. It's for developers adding transcription to applications and people who need offline speech-to-text.

Strix is an AI penetration testing tool for developers and security teams assessing code, web applications and APIs. Its free, open source edition runs locally with Docker under the Apache 2.0 license. It supports local models through Ollama or LM Studio, as well as remote providers such as OpenAI and Anthropic; cloud-backed runs send model requests to the selected provider.

Libre WebUI is a self-hosted AI workspace for individuals and teams who want chat, document research and AI agent tasks on infrastructure they control. You can use local models or connect a remote provider of your choice.

BentoML is a Python framework for developers who want to turn AI models into services on their own hardware or servers. It handles model serving and application logic for LLMs, image generation, audio, and embeddings.

Wyoming Piper provides local speech generation for Home Assistant and other Wyoming-compatible clients. It's for people building a voice assistant who want speech synthesis to run on their own machine or server.

Cognee gives AI agents memory across sessions so they can reuse project context, decisions, and lessons from earlier work. It's for developers building agents and teams whose knowledge sits across documents, code, and conversations.

ZeroClaw is a personal AI agent for people who want an assistant running on their own computer or server. It can use a shell and browser, respond through messaging apps and perform scheduled tasks.

Open Generative AI is an AI image and video studio for creators who want to generate media on their own hardware. Local inference is available only in the desktop app.

OpenVoiceOS is a voice assistant platform for people building smart speakers or adding voice control to their own devices. Its core is a fork of MycroftAI/mycroft-core, and most classic Mycroft skills remain compatible.

FastSurfer processes structural human brain MRI scans on your own workstation or server, producing brain segmentations and cortical measurements compatible with FreeSurfer. It's for neuroimaging researchers working with individual scans or large cohorts who want shorter processing times within existing analysis workflows. The software is open source under Apache 2.0.

SQL Chat is a SQL client for people who want to work with databases using natural language. Its chat interface lets you ask for data and request database changes without writing every SQL statement yourself.

TabbyAPI is a self-hosted LLM API server for people who want to run ExLlamaV3 models on their own hardware. It’s built for personal projects and small groups; its maintainers don’t recommend it for production servers.

Spleeter is Deezer's music separation library for developers and audio researchers who want to extract individual parts from a recording on their own hardware. Its pretrained models let you process music without training a model first.

Revornix is an AI knowledge workspace for people who collect articles, research papers and recordings and want to ask questions across their library. It connects saved material in a personal knowledge graph and uses those relationships to inform its answers.

Nexent is a self-hosted AI agent platform for people and teams who want to build agents by describing tasks in plain language. It generates executable agents without requiring code or a visual workflow editor.

SapienAI is an academic chatbot and research workspace for researchers who want to keep source material, writing and AI conversations together. Research spaces group project documents, notes and drafts so the AI can answer questions using uploaded material.

RAG Web UI is a self-hosted app for asking questions about your own documents. It's a learning project, and its maintainers advise against commercial use and state that it isn't ready for production.