OpenHands

https://openhands.dev

$0/month (free, open source; bring your own LLM API key)

Autonomous Coding Agent — Plans, writes, tests, and fixes code end-to-end without continuous supervision
Docker Sandbox — Every action runs in isolated containers; safe to let the agent work unattended
Model-Agnostic (LiteLLM) — Works with Claude, GPT-4o, DeepSeek, Qwen, Gemini, and any OpenAI-compatible API
Three Interfaces — CLI (headless), web GUI (local Docker), managed cloud (GitHub integration)
Pre-Built Workflows — Fix CVEs, review PRs, migrate code, triage incidents
SWE-bench Verified — State-of-the-art performance on the coding benchmark that matters
GitHub Integration — Tag the agent on an issue or PR and it opens a branch automatically
Software Agent SDK 1.0 — Modular SDK for building custom AI agents on top of OpenHands

Top benchmark performance on SWE-bench Verified
Fully open source (MIT license) — no vendor lock-in
Model-agnostic: works with Claude, GPT-4o, DeepSeek, Qwen, Gemini
Isolated Docker sandbox — safe to run unattended
Pre-built workflows: vulnerability fixes, PR reviews, code migrations
GitHub integration for automated PR/issue workflows
New Software Agent SDK (1.0) for building custom agents
44K+ GitHub stars, active community

Requires LLM API key setup — not turnkey out of the box
GUI is functional but not polished for non-technical users
Task decomposition can drift on complex multi-repo scenarios
No built-in model — you pay your own LLM costs
Cloud service sends code to third-party runtime