helix
♾️ Private Agent Fleet with Spec Coding. Each agent gets their own GPU-accelerated desktop. Run Claude, Codex, Gemini and open models on a full private AI Stack ♾️
Helix is a self-hosted platform for running fleets of autonomous coding agents on private infrastructure, where each agent operates in its own GPU-accelerated desktop environment. It enables teams to manage multiple AI coding agents in parallel through a spec-driven Kanban board, supporting agents like Claude Code, Codex, Gemini CLI, and Qwen Code. The platform provides full isolation for each agent with browser, terminal, and filesystem access, while allowing teams to collaboratively dispatch tasks, review pull requests, and switch between different LLM providers including self-hosted models. Helix runs entirely on your own infrastructure with complete air-gap capability.
Key Features
Use Cases
- 01Running parallel coding agents across multiple repositories with each agent in isolated sandboxes
- 02Managing software development tasks through AI-generated specs that require human review before implementation
- 03Operating AI coding assistants in air-gapped or high-security environments with fully self-hosted infrastructure
- 04Reducing LLM costs by starting tasks on local models and escalating to frontier models only when needed
- 05Enabling distributed teams to collaborate on AI-assisted development with shared agent environments and persistent state
- 06Implementing automated code generation workflows with pull request review gates integrated into existing Git workflows
Related Agents
View morehermes-agent
The agent that grows with you
agency-agents
A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables.
openinterpreter
A coding agent for open models like Kimi K3
cline
Autonomous coding agent as an SDK, IDE extension, or CLI assistant.
helix — FAQ
What is Helix and what does it do?+
Helix is a self-hosted platform for running fleets of AI coding agents, where each agent gets its own isolated desktop environment with GPU acceleration. It manages agent tasks through a Kanban board and automatically creates pull requests in your repositories.
How do I install Helix on Docker?+
Download and run the quickstart installer script with curl, which will prompt you before making system changes and set up Helix on http://localhost:8080 by default. For production deployments with DNS and TLS, use the Kubernetes Helm charts instead.
Which agent harnesses work with Helix?+
Helix supports Claude Code, OpenAI Codex, Gemini CLI, Qwen Code, Goose, Zed Agent, and any ACP (Agent Client Protocol) compatible agent. You can switch between different harnesses mid-task.
What API keys or prerequisites do I need?+
You need API keys for your chosen LLM provider (OPENAI_API_KEY, ANTHROPIC_API_KEY, etc.), PostgreSQL database connection details, and GPU runners if you want to use self-hosted models. For local development, you need Docker, Go 1.24.0+, and Node.js 18+.
Is Helix free to use?+
Helix is free for personal use, educational use, and small businesses with under $10M annual revenue and less than 250 employees. Larger commercial organizations need to purchase a license through the Launchpad, with trial licenses available.
Can I run Helix with self-hosted LLMs?+
Yes, Helix supports any OpenAI-compatible endpoint including vLLM, allowing you to run open models on your own GPUs, on Kubernetes or bare metal, including fully air-gapped deployments.
How do I install helix?+
Open the source repository on GitHub and follow its README. helix is a agent — MCP Agents Market links you directly to the official repo.
Is helix free?+
helix is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.