What Are CPU Seconds?
CPU seconds are how Latenode measures compute time when your scenario runs. They help you understand what each execution costs and how much free usage you have left. On the Free plan you get 10,000 CPU seconds per month; after that, Pay as You Go rates apply. Open Plans & Pricing in the left menu to see your balance, or check Usage Statistics for per-run details.
How CPU seconds are charged
- CPU seconds are consumed during a scenario execution (from start to finish of the run).
- Billing is based on total compute time for the run, not on the number of nodes or operations.
- On the Free plan, usage is billed in 1-second steps. Minimum billable execution time is 1 CPU second per run.
The number of nodes does not directly change the rate. A scenario with 10 nodes and one with 2 nodes that both finish in a few seconds are each charged at least 1 CPU second on Free. Scenarios with more nodes or heavier data processing usually take longer and use more CPU seconds.
Running a single node does not consume CPU seconds, so you can test and configure nodes on the canvas for free. When you run a full scenario that includes a nested Latenode scenario node, CPU seconds are counted for the main scenario's execution time only.
Free tier and Pay as You Go
| CPU seconds (per month) | Cost |
|---|---|
| 0 - 10,000 | Free |
| 10,001 - 100,000 | $0.00012 per CPU second |
| 100,001 - 200,000 | $0.00011 per CPU second |
| 200,001 - 300,000 | $0.00010 per CPU second |
| 300,001 - 400,000 | $0.00009 per CPU second |
| 400,001 - 500,000 | $0.00008 per CPU second |
| 500,001 - 1,000,000 | $0.00007 per CPU second |
| 1,000,001 - 5,000,000 | $0.00006 per CPU second |
| 5,000,001 and above | $0.00005 per CPU second |
The more CPU seconds you use in a billing period, the lower the per-second rate in your current tier. Full plan details, add-ons, and payment setup are on the Billing page.
Duration vs CPU seconds in Execution History
Execution History shows two time-related columns:
| Column | What it measures |
|---|---|
| Dur. (Duration) | Wall-clock time of the run as shown in the UI |
| CPU seconds | Compute time billed for the run |
These numbers are often not the same. Queue time, platform overhead, and other factors can make Duration and CPU seconds differ. Use the CPU seconds column (and Usage Statistics) for billing; use Duration to see how long the run took from your perspective.
Runs blocked by spending limit
If you set a spending limit for CPU seconds on the Billing page and reach it, new scenario runs fail before they start. Raise or remove the limit under Plans & Pricing, or wait until the next billing period if usage resets.
Plug&Play Tokens are separate
PnP nodes (built-in AI models and similar integrations) bill Plug&Play Tokens on top of CPU seconds. One PnP token costs $1.00. See Plug-n-Play (PnP) Tokens for details.
Legacy accounts
Some older accounts still see credits instead of CPU seconds in the UI. On those accounts, 1 credit equals 30 CPU seconds. For the full legacy guide (Boosts, Extra Credits, minimum charges by plan), see Legacy Execution Credits.
How to get Support
How to get Support
Plug-n-Play (PnP) Tokens
What PnP tokens are and how $ nodes are billed.
Need Help? Ask the community
If something on this page is missing or unclear, post on the Latenode community forum. Our team and other users usually reply quickly.