MAESTRO is a self-hosted AI research assistant that manages complex research tasks from start to finish: you plan a research mission, AI agents carry it out across your documents and the web, and you receive detailed, sourced reports. It is open source on GitHub with about 1,500 stars, ships as a Docker-ready application, and saw continued releases and repository activity into 2026.
Deep research tools are usually locked inside someone else’s cloud. MAESTRO is the version you host on your own hardware. You build a document library — grouped and organized the way you work — and define research missions with your own settings. Agents then read your sources alongside material from the web, and generate structured reports you can inspect and refine. A chat interface lets you question your document collection directly, and a writing assistant helps turn findings into finished text. MAESTRO works with multiple AI providers, including Azure OpenAI with support for GPT-5-class models, and lets you enter model details manually for providers that need custom configuration.
MAESTRO is free and open source under the AGPL-3.0 license. You host it yourself, so the only ongoing costs are your own infrastructure and whatever model API usage your research missions consume through your chosen provider.
Analysts, researchers, students, and small teams who want deep-research capabilities — long, sourced, multi-document reports — without sending sensitive documents to a third-party cloud service, and who are comfortable running a Docker application.