Pricing with Hard Caps
Predictable costs for the Web OS for AI Agents. Every plan has hard billing caps — you know your max spend, no surprises, even with 24/7 scheduled agents or complex multi-agent workflows.
No credit card. No API keys. Hosted frontier models included.
Hard Caps on Every Plan
Launch free and scale as your agent teams grow. Hard caps mean you can run monitoring agents 24/7, give research agents long context, and experiment with workflows without worrying about overages.
Free
Start here. Full OS access with generous included credits and a hard cap.
No credit card. Cancel anytime.
Growth / Pro / Max
Higher hard caps + priority for production agent teams. Estimate exact needs with our free tools before upgrading.
All plans include hosted models (Claude, GPT-4o/o1, DeepSeek, Moonshot + more), real cloud browser sessions, persistent workspace, multi-agent collaboration, scheduling, and connections — with hard caps.
Free Tools to Model Your Spend
Before you scale, use these client-only tools (no data sent) to estimate tokens, model mixes, browser/tool use, and scheduled runs.
See the dedicated blog post for details: What It Costs to Run AI Agents: Tokens, Models, and Predictable Caps.
Launch with confidence
Hard caps + visibility in the OS desktop = predictable AI agent operations from day one.
Launch CloudAxis OS — free