Real Pricing Compared: Tasks vs Operations vs Executions
· By AI Compare
A simple formula and three worked examples to compare automation costs across Zapier, Make and n8n.
Why automation pricing is confusing
Each tool bills a different unit:
| Tool | Unit | What counts |
|---|---|---|
| Zapier | Task | Every successful action step. Triggers and filters that stop a Zap do not count. |
| Make | Credit (formerly operation) | Every module run, including the trigger and each iteration in a loop. AI modules can use more. |
| n8n Cloud | Execution | One full workflow run, regardless of steps. |
| n8n self-hosted | None | Unlimited; you pay for the server. |
So "750 tasks" vs "10,000 credits" vs "2,500 executions" are not comparable until you apply them to your own workflow.
The calculator formula
Work out three numbers for each workflow:
- R = runs per month
- A = action steps per run (excluding the trigger)
- L = average items processed per run inside loops (1 if no loops)
Then:
- Zapier tasks ≈ R × A × L
- Make credits ≈ R × (1 + A × L), plus 1 credit for each scheduled check
- n8n Cloud executions ≈ R
- n8n self-hosted = server cost only
Add the totals for all your workflows and match them to each plan.
Interactive calculator coming soon: enter your workflows and see monthly cost in all three tools.
Worked examples
Example 1: small business
A form-to-CRM workflow: trigger + 3 actions, 300 runs a month.
| Usage | Plan needed | Approx. cost | |
|---|---|---|---|
| Zapier | 900 tasks | Professional, ~1,000-task tier | ~$25–30/mo |
| Make | ~1,200 credits | Core | ~$10–12/mo |
| n8n Cloud | 300 executions | Starter | €20/mo |
| n8n self-hosted | – | – | $6–10/mo server |
At this size, Make is cheapest, and the difference is small. Choose on ease of use.
Example 2: growing team
Six workflows averaging trigger + 5 actions, 3,000 runs a month in total.
| Usage | Plan needed | Approx. cost | |
|---|---|---|---|
| Zapier | 15,000 tasks | Higher Professional or Team tier | Typically $150+/mo |
| Make | ~18,000 credits | Core with a larger credit pack | Typically $20–35/mo |
| n8n Cloud | 3,000 executions | Pro (Starter covers 2,500) | €50/mo |
| n8n self-hosted | – | – | $6–10/mo server |
Example 3: batch processing
A nightly sync that processes 500 records through 4 steps, every day.
| Usage | Approx. cost | |
|---|---|---|
| Zapier | ~60,000 tasks | Several hundred dollars a month |
| Make | ~60,000 credits | Large credit pack, often $100+/mo |
| n8n Cloud | 30 executions | Starter, €20/mo |
| n8n self-hosted | – | Server only |
This is where n8n's per-execution model wins by an order of magnitude.
Approximate costs based on public pricing in October 2026. Zapier and Make price larger tiers on sliders; check their pricing pages for exact figures.
Hidden costs to include
- Self-hosting time: updates, backups and monitoring. Budget a few hours a month or a managed host.
- Premium apps: some Zapier integrations require higher plans.
- AI usage: LLM calls cost tokens on top of automation pricing, regardless of tool.
- Polling: in Make, scheduled checks that find nothing still consume a credit each; frequent schedules add up.
- Migration: rebuilding workflows takes time; see our migration guide.
How to cut costs in any tool
- Filter as early as possible so unwanted records stop before expensive steps.
- Use instant webhooks instead of frequent polling.
- Batch: process many records in one run instead of one run per record.
- Combine several small automations into one where it makes sense.
