Pricing
Pay only when work executes
No reserved capacity or minimum commitments
Pricing scales with usage, not speculation
Routing decisions optimize both performance and cost
You are billed for useful work. Not for idle resources, pre-provisioned instances, or
infrastructure assumptions.
Work executes
Function, agent, or workflow runs
Usage measured
Runtime and compute tracked
Cost applied
Pay for what ran
What you pay for
Execution usage
Metered by execution time and workload requirements. Applies to functions, agents, and workflows. No billing for idle resources.
Marketplace usage
Functions are priced per execution by builders. Costs are visible before workloads run and scale with actual usage.
Platform services
Execution routing, monitoring, and platform management required to run workloads reliably at scale.
What you do not pay for
Enterprise
Designed for scale and
governance
Enterprises running high-volume or sensitive workloads can access tailored pricing, deployment options, and governance controls aligned to their execution needs.
Volume-based pricing adjustments
Deployment flexibility across cloud, hybrid, or on-prem
Security and audit capabilities based on workload requirements
Priority onboarding and support
As workloads stabilize, execution patterns become more predictable and pricing aligns closely with real output.
Usage visibility
Request volume
12.4k calls
Execution time
847 min
Cost accumulation
Tracked in real-time
