OpenHands
Run coding agents and engineering automations through Agent Canvas with local, Docker, remote or cloud backends.
These projects connect a model to tools for a specific kind of work: research, coding or a longer task with files and a browser. The repository contains the software; you still need a compatible model service or local model, a place to run it and permission to use the connected tools.
This is a selected reading list, not a popularity ranking. Each profile explains the current project, its setup and operating costs, the access it needs and what the published examples actually demonstrate. Follow the implementation links to see an output or documented build before planning your own trial.
Open-source licensing covers the project code, not every model, hosted service or connected account. Local execution can still send prompts and files to a cloud model. Cost and maintenance evidence are dated in each profile.
Run coding agents and engineering automations through Agent Canvas with local, Docker, remote or cloud backends.
Generate cited research reports from web sources or supplied documents using planning, retrieval and report-writing stages.
Build a configurable agent with planning, files, subagents, skills and context management already assembled on LangGraph.
Run longer research and artifact-creation tasks using a lead agent, delegated agents, skills, memory and execution sandboxes.
Use a native desktop, CLI or API agent with a chosen model and extensions for coding, research and everyday computer work.