Pricing that scales with you.
Free for individuals. Pro for builders. Team and Enterprise for organizations that need security and scale.
For individual developers exploring XR.
Get started- XR CLI + dashboard
- Unlimited local runs
- Access to free models
- Community skills
- Public marketplace
- Community support
For developers who ship with XR daily.
Start Pro trial- Everything in Free
- Pro models (GPT-5, Claude Opus, XR Core)
- Priority skill access
- Private skills & teams
- Advanced MCP servers
- Session replay & trace search
- Priority support
For teams collaborating on agents and workflows.
Start Team trial- Everything in Pro
- SSO / SAML
- Shared workspaces
- Role-based access
- Audit logs (90 days)
- Custom model gateways
- SLAs & dedicated support
For organizations that need security, scale, and control.
Contact sales- Everything in Team
- Self-hosted runtime
- VPC deployment
- Dedicated models
- Custom skills & SLA
- SOC 2 / HIPAA / ISO 27001
- Dedicated solutions engineer
Frequently asked
Is XR open-source?+
Yes. The XR runtime, CLI, and most official skills are MIT-licensed. Some cloud-only features ship under a commercial license.
Which editors are supported?+
VS Code, Neovim, JetBrains IDEs, Zed, Cursor, and Windsurf all have first-party extensions. The CLI works anywhere.
Can I run XR locally?+
Absolutely. The runtime is designed for local-first operation with Ollama, Llama.cpp, vLLM, and any OpenAI-compatible endpoint.
What is a skill?+
A skill is a typed, versioned, capability-scoped unit of work — think of it as a better tool. Skills compose into agents.
How is XR different from coding agents?+
Coding agents are one feature of XR. XR is a full runtime: shell, editor, skills, models, security, and deployment in one coherent system.
Do you support MCP?+
Yes. XR is MCP-native — you can use any MCP server without an adapter. We also extend MCP with typed streams and state.