Plan tiers
These are representative limits. Check your actual limits with the CLI - they may differ based on promotions or custom arrangements.
Check your limits
Show your current plan limits and usage:Workspace resources
When creating a workspace, you specify RAM, vCPU, and disk allocation. These are validated against your plan limits:402 Payment Required error indicating which limit was exceeded.
Resizing
You can resize a workspace after creation. The workspace must be stopped first:AI credits
AI credits fund requests through the managed AI proxy. Each credit maps to approximately $0.01 of upstream provider cost, though the exact rate depends on the model and token count.Check your balance
View usage breakdown
When credits run out
- AI requests through the proxy return
402 - Non-AI services in your workspace continue running
- Switch to BYOK mode for unlimited usage with your own keys
Rate limits
API requests are rate-limited per user. Limits vary by endpoint class:
When rate-limited, the API returns
429 Too Many Requests. Implement exponential backoff in your client:
Capacity
Before creating a workspace, check if the platform has capacity:172.16.0.2 through 172.16.0.254), and stopped or deleted workspaces release their active IP lease.
Learn more
Architecture
How the platform is structured
Managed AI Proxy
How credits and the AI proxy work