Latenode vs n8n: compute seconds vs executions, costed side by side

Compute time vs number of runs

Latenode bills “CPU seconds”: the compute time of each scenario execution, from start to end. n8n bills “executions”: one complete workflow run counts once. Neither counts steps. If your runs are short, Latenode’s meter barely moves; if they are long, n8n’s does not move any faster.

Three more gaps weigh on the choice. Latenode has a permanent free plan; n8n Cloud has only a 14-day free trial. n8n can be installed on your own server; for Latenode, that point was not recorded. And the two price scales were not recorded in the same currency (US dollars for Latenode, euros for n8n). This page analyzes both vendors’ public pricing and documentation, recorded on the date shown.

The same workflow on both

Take a lead form on your website. Each entry arrives by webhook (an instant trigger) and sets off four actions: add a row to a spreadsheet, create a contact in the CRM, send an email, post a message in a team chat. Assumption: each run takes 5 seconds. That is the default in the cost calculator; replace it with your own figure.

At 1,000 entries a month.

• Latenode: 1,000 × 5 seconds = 5,000 CPU seconds. The Free plan includes 10,000 a month, so the workflow fits.
• n8n: 1,000 executions. The Starter plan includes 2,500 executions a month for €24 a month.

At 10,000 entries a month.

• Latenode: 50,000 CPU seconds. You have to turn on “Pay as you go”: the first 10,000 seconds stay free and the next 40,000 are billed per second, at the rate of the first paid bracket.
• n8n: 10,000 executions, exactly the first tier of the Pro plan (€60 a month), with no headroom.

The number of actions changes nothing on either side. What changes everything is run time:

Monthly volume	Latenode	n8n
1,000 runs of 5 seconds	5,000 CPU seconds (Free)	1,000 executions (Starter)
1,000 runs of 60 seconds	60,000 CPU seconds, 50,000 of them billed	1,000 executions (Starter)
10,000 runs of 5 seconds	50,000 CPU seconds, 40,000 of them billed	10,000 executions (Pro, 10,000 tier)
100,000 runs of 1 second	100,000 CPU seconds, 90,000 of them billed	100,000 executions: beyond Pro’s last tier (“Custom plan”)

Both price scales

The Latenode box gives the rate for each bracket; the table after it repeats the vendor’s own worked examples.

Latenode volume tiers

Plan	Monthly volume (CPU seconds)	Monthly price	Monthly price, billed annually
Free	10,000	$0	not recorded
Pay as you go	100,000	$10.80	not recorded
Pay as you go	500,000	$48.80	not recorded
Pay as you go	1,000,000	$83.80	not recorded

n8n volume tiers

Plan	Monthly volume (executions)	Monthly price	Monthly price, billed annually
Starter	2,500	€24	€20
Pro	10,000	€60	€50
Pro	50,000	€144	€120
Business	40,000	€800	€667
Business	50,000	€985	€821
Business	75,000	€1,456	€1,213
Business	100,000	€1,726	€1,438
Business	150,000	€2,189	€1,824
Business	200,000	€2,639	€2,199
Business	300,000	€3,512	€2,927
Business	400,000	€4,381	€3,651
Business	500,000	€5,245	€4,371

In the n8n table, the Business rows belong to a self-hosted-only plan. The n8n pricing page goes through each plan.

What makes the bill grow

On Latenode, anything that makes a run longer. The vendor says so itself: two scenarios with the same number of executions can cost different amounts, because waiting on slow APIs, large files and the headless browser lengthen the billed time. On the Free plan, every execution counts for at least 1 second. The documentation also separates, in the run history, elapsed time (“Duration”) from billed time (“CPU seconds”): the second column is the one to read on your own scenarios before doing the math (Latenode documentation, FAQ on the billing unit (accessed November 11, 2026)).

On Latenode again, add-ons and AI. The minimum interval for scheduled or polling triggers is 10 minutes; 5 minutes or 1 minute are paid add-ons, as is a longer maximum run time. AI models used without an API key are billed in “PnP tokens,” on top of CPU seconds.

On n8n, scheduled triggers. A Schedule Trigger counts one execution each time it fires: a workflow scheduled every 15 minutes produces 2,880 executions a month, more than Starter’s quota. A webhook counts one execution per incoming request, even with an empty body.

Past the quota. On Latenode, a monthly spending cap can be set: once it is reached, new executions fail before starting. On n8n, Starter and Pro are “Capped”: according to the tooltip on the pricing page, workflows keep running and these two plans have no overage fees.

The guide to tasks, credits, operations and executions sets these two rules next to those of four other tools. Against Make, which counts every module that runs, the reasoning about run time is the same: see also n8n vs Make.

Self-hosting and where your data lives

n8n publishes a free Community edition, with no license key, that you install on your own server; the vendor recommends Docker and reserves self-hosting for experienced users. Its “fair-code” license allows use for your own internal business purposes. See n8n Cloud vs self-hosted. For Latenode, neither self-hosting nor a license was recorded.

On data, the gap is clear. n8n’s hosted plans store data in the EU, on servers in Frankfurt. For Latenode, the answer recorded is: Hosting on Microsoft Azure and Amazon Web Services, both listed as "US" in the list of subprocessors source.

Code, browser and AI

• Latenode: the JavaScript node and the headless browser are included in both plans, Free included. There is an “AI Agent” node; models used without an API key are paid for in “PnP tokens.”
• n8n: the AI Agent node is included on Cloud and in the Community edition; model calls are paid with your own provider keys or with Gateway credits. The MCP protocol is included in every Cloud plan and in the Community edition. n8n’s code step was not recorded in our database; MCP was not recorded for Latenode.

What does not separate them, and what is left

• The number of integrations: no clean figure. Latenode shows “5,500+ Integrations” and, on its “About” page, “1,200+ apps and LLMs in one plan.” On n8n, the pages contradict each other: “500+ native integrations” on the home page, “2,000+ integrations” on the integrations page.
• Active scenarios: 5 on Latenode’s Free plan, unlimited on “Pay as you go”; unlimited on every n8n plan.
• Maximum run time: 3 minutes (Free) and 10 minutes (“Pay as you go”) on Latenode; 5 minutes (Starter) and 40 minutes (Pro) on n8n.
• What is missing: support options were not recorded for Latenode.

For other options, see the n8n alternatives and the comparison of automation tools.