We earn commissions when you shop through the links below, at no extra cost to you. It never changes our rankings or numbers. How this site earns.
What counts as a task?
The same workflow uses very different amounts of each platform's allowance. Here is each one's billing unit, in its own words.
One workflow, six counts
A workflow that runs 1,000 times a month, has 4 action steps, and checks for new data every 15 minutes (2,880 checks a month) uses:
| Platform | Units a month | Unit |
|---|---|---|
| Zapier | 4,000 | tasks |
| Make | 6,880 | credits |
| n8n Cloud | 1,000 | executions |
| Activepieces | 1,000 | runs |
| Pabbly Connect | 4,000 | tasks |
| Integrately | 4,000 | tasks |
Units aren't comparable on their own: what each costs depends on the plan. Put your own workflow in the calculator to see the monthly price.
Zapier: tasks
- What counts
- Each action step counts. The trigger is free. Triggers, polling checks and built-in tools like Filter and Formatter are free.
- In its words
- "When Zapier performs an action successfully, it counts as a task. Each successful action in a Zap counts as a separate task."
Zapier's other rules
What it doesn't count, polling and what happens past the plan, in its words
- Not counted
- “Formatter, Paths, Filters, Delay, Looping, Sub-Zaps, Digests, Zapier Manager, Storage, Tables, and Forms; these steps never use tasks.”
- Polling
- Zapier checks every 15 minutes on Free, every 2 minutes on Professional, every minute on Team (“Polling time 15 min 2 min 1 min 1 min”; on what a check costs, “Zapier never charges you a task to check for new data—only when a Zap successfully completes an action for you.”).
- Past the plan
- “If you exceed your plan's task limit, you'll automatically switch to pay-per-task billing for the remainder of your pay period, so your Zap workflows stay on.”, “Pay-per-task billing is charged at a higher per-task rate than your base subscription tasks.”, “Your pay-per-task rate depends on your plan and billing cycle type. Check your Billing settings for the exact rate.” and “Zap workflows will stop running if your account reaches the maximum pay-per-task usage amount. This is set at 3 times your plan's task limit, which includes your plan allocation and optional pay-per-task usage.”
Source: Zapier pricing; polling and overage terms as its pricing page, help center or docs put them.
Make: credits
- What counts
- Each module counts, the trigger included. A polling trigger uses one on every check, even when nothing new arrived. The trigger counts too, and a polling trigger uses a credit on every check, even when it finds nothing.
- In its words
- "Each module action in your scenario, like adding a Google Sheet row or fetching Gmail account data, counts as one credit."
Make's other rules
What it doesn't count, polling and what happens past the plan, in its words
- Not counted
- “Router module - no credits counted”
- Polling
- Make checks every 15 minutes on Free, every minute on Core, Pro and Teams (“Minimum time (in minutes) between scheduled scenario executions. 15 1 1 1 1”; on what a check costs, “The Google Sheet > Watch New Rows module checks for new rows in a Google spreadsheet = 1 operation (one per check)”).
- Past the plan
- “Manually purchased extra credits include a 25% additional cost compared to the credits in your plan. The cost of extra credits is the same whether you purchase them manually or through auto-purchasing.” and “When you use all of the credits in your subscription, Make automatically purchases 10,000 extra credits. All automatic purchases are in units of 10,000 credits. Make then bills you for the 10,000 extra credits.”
Source: Make pricing; polling and overage terms as its pricing page, help center or docs put them.
n8n Cloud: executions
- What counts
- One per run, however many steps the workflow has. One execution per run, however many steps. Self-hosting the Community Edition is free apart from your server.
- In its words
- "An execution is a single run of your entire workflow. It doesn't matter how many steps are in the workflow or how much data it processes. It's still a single execution."
n8n Cloud's other rules
Polling and what happens past the plan, in its words
- Polling
- n8n Cloud publishes no polling interval (on what a check costs, “Polling nodes (like Google Drive Trigger): Count one execution only when new data is found. Polls that return no results don't count as an execution.”).
- Past the plan
- its cloud plans publish no price for going past them; for Business and Enterprise it answers “If you exceed your quota, rest assured your workflows will continue running without interruption, but overage charges may apply if you do not upgrade to the next usage tier.”
Source: n8n Cloud pricing; polling and overage terms as its pricing page, help center or docs put them.
Activepieces: runs
- What counts
- One per run, however many steps the workflow has. One credit per run; AI steps cost extra credits.
- In its words
- "A run is one credit however many steps it has, and AI is the exception that sits on top, because we are paying the model provider on your behalf."
Activepieces' other rules
Polling and what happens past the plan, in its words
- Polling
- Activepieces checks every 5 minutes on Plus and Team (“This method runs every 5 minutes”).
- Past the plan
- “On Plus and Team they keep running and you pay for the extra at $0.007 per credit.”
Source: Activepieces pricing; polling and overage terms as its pricing page, help center or docs put them.
Pabbly Connect: tasks
- What counts
- Each action step counts. The trigger is free. Triggers and internal steps (filter, router, formatter) are free.
- In its words
- "Any action performed within your workflow is considered a task. Triggers are not included in the task count. Only action steps are included. Internal tasks of Pabbly Connect, such as Filter, Router, and Formatter, are not counted as tasks."
Pabbly Connect's other rules
Polling and what happens past the plan, in its words
- Polling
- Pabbly Connect publishes no polling interval (on what a check costs, “So, if a check is made every 5 minutes, that is 288 operations in a day, whether or not there is any new data. However, Pabbly Connect does not charge for operations.”).
- Past the plan
- “You can also stack or purchase multiple Standard Plans to increase your task limit.”
Source: Pabbly Connect pricing; polling and overage terms as its pricing page, help center or docs put them.
Integrately: tasks
- What counts
- Each action step counts. The trigger is free. Conditions, searches and data edits count as tasks too, so real use is often higher than shown.
- In its words
- "Every successful execution of action, condition / branching, modify data (inline in an action or independent step), search (inline in an action or independent step) is counted as a task."
Integrately's other rules
What it doesn't count, polling and what happens past the plan, in its words
- Not counted
- “Triggers are free. Failed actions are also NOT counted.”
- Polling
- Integrately checks every 15 minutes on Free, every 5 minutes on Starter, every 2 minutes on Professional, Growth and Business (on what a check costs, “Triggers are free. Failed actions are also NOT counted.”).
- Past the plan
- “Don’t worry, your automations won’t stop. But your actions won’t execute.” and “And once you upgrade / are allotted tasks, you will be able to resume the execution of actions from the History page.”
Source: Integrately pricing; polling and overage terms as its pricing page, help center or docs put them.
Platforms that don't charge per run
- Microsoft Power Automate
- Sells a licence per user or per flow, with a daily cap on actions (40,000 a day on Premium per user) rather than a monthly allowance (Power Automate pricing).
- IFTTT
- Sells a monthly allowance of credits that Applet runs, MCP use and AI features spend, with unlimited Applets on every plan; its pricing page doesn't say how many credits a run uses (IFTTT pricing).
Platforms that charge credits for agent work or compute
These sell credits that don't map to a run or a step, so the table above can't count them:
- Gumloop
- AI agents: A credit is $0.005 of the agent's cost: model tokens at the provider's list price, 1 credit per connector tool call, 5 credits per minute of active compute, plus an 8% orchestration fee; what a run uses varies. In its words: “Agent credit costs are variable. The same agent might cost 2 credits for a quick question and 200 for a deep research task.” (Gumloop pricing)
- Lindy
- AI teammate: Billed per user a month; every seat adds its credits to one shared pool, and Lindy puts a job at 2 to 2,500 credits depending on its size. In its words: “Credits measure the work Lindy does. You only use them when Lindy is working.” (Lindy AI pricing)
- Pipedream Connect
- Integration API for developers: 1 credit per 30 seconds of compute for each action run, MCP tool call, proxy request or deployed-trigger event; past 100 external users, $2 a user a month. In its words: “Credits measure Connect API usage. Running an action, calling a tool through MCP, sending a request through the Connect proxy, and emitting an event from a deployed trigger each use 1 credit per 30 seconds of compute.” (Pipedream pricing)
Prices checked against each platform's official pricing page on .