Usage and Credits

Atlasflow bills for the resources your deployments actually use, metered by the second. Your plan's monthly credits are applied first, so you only pay more if your usage exceeds your plan minimum.

# Usage rates

The same per-second rates apply on every plan:

Usage is prorated down to the second. There's no per-request pricing and no cold-start penalty, because your workload runs on a dedicated microVM rather than a shared pool.

# How credits work

Each plan includes a monthly credit allowance equal to its minimum:

Credits are granted at the start of each billing period and applied to that period's usage. They expire at the end of the period, so unused credits don't roll over. When your usage in a period exceeds your credits, the overage is billed.

# What you're billed for

• Compute (vCPU) and memory are billed only while your instances actually run, prorated to the second. An idle, scaled-down, or stopped instance costs less.
• Storage and traffic are billed for what you use over the period.

The metrics on your deployments reflect the same CPU and memory usage that feeds billing, which makes them a good way to predict your bill.

# Estimating your bill

Use the interactive estimator on the pricing page to model a workload like yours. Because billing is per-second, real bills are typically lower than a worst-case estimate that assumes constant full utilization.

# Next steps

• Plans: compare what each plan includes.
• Managing billing: invoices and payment.
• Metrics: watch the usage that drives your bill.