StackReceipt

Automation pricing

What does your
workflow use?

Translate your workload into tasks, credits, and executions. Check a focused set of base-plan allowances and dated fees before requesting comparable quotes.Compare tasks, credits, and executions against checked base-plan allowances. Fees and uncertainty stay visible.

A workflow, counted three ways
  1. TriggerNew lead arrives
  2. Action 1Add to CRM
  3. Action 2Notify the team
Zapier2 action tasksTrigger charges unresolved
Make3 operationsTrigger + two actions
n8n1 executionThe whole workflow
Modeled example: one lead, two ordinary actions, no retries or extra checks. Zapier’s official sources conflict on trigger charges; two action tasks exclude any trigger task. Make uses three credits here only at its standard one-credit module rate. See assumptions

Use-case assistant

Not sure? Describe your workload.

Use this fallback when a preset does not fit. Explain what runs, how often, and whether it polls for updates; the calculator remains available without it.

Going live soonThe assistant is being connected. The calculator and guides below are available now; recommendations and answers will come from the verified pricing API once it is live.

Cost calculator

Compare your workload

Enter your workload. See usage units and checked plan allowances.

Your workload

Start with an example, then tell us how much work you handle. We’ll do the monthly arithmetic.

Dated price observations, not live quotes. Source dates appear with each result.

Describe your workflow

Describe the job in everyday language. We’ll count the steps and show a starting estimate. If a detail is missing, we’ll ask one question at a time.

Your description stays in this browser and is not included in shared links.

Use an example or enter numbers yourself

Choose a workflow

Pick a starting point, then adjust the numbers to match your work.

Choose a workflow to see its starting assumptions.

tier

1 Solo · 2 Startup · 3 Mid-market · 4 Enterprise

Startup

Make the example yours

Example size only changes the starting numbers. It is not a vendor pricing tier.

Count each candidate, lead, ticket, or other item once. We assume one workflow starts for each item.

For example: save a record + send an email + notify a teammate = 3 actions. Don’t count the event that starts the workflow.

Help me count the actions

Select the actions in your workflow. Each selection counts once; add repeated or other actions using the number above.

Help me choose a tool, not just compare cost

Optional documented examples. These verify components, not account permissions or end-to-end setup. Selecting one does not change your entered volume or action count.

Optional. We’ll include these in your evaluation checklist. We have not verified their exact triggers, actions, or plan requirements. This answer stays on this page and is not added to shared links.

This is a starting estimate. Scheduled checks, branches, and batches can change usage; adjust these under Advanced inputs.

Or use a workload preset

These existing scenarios load fixed totals independently of the use-case model.

workflows

How many active automations contribute to these totals. Make Free supports up to two active scenarios; this input checks that limit.

actions

Each step performed after a trigger.

runs

Successful items processed across your workflows. Batch size converts these into workflow starts.

steps

Each action performed after the trigger.

Leave at 1 for event-by-event work. A batch of 10 items shares one trigger, but downstream actions still run per item. Branches, retries, and weighted AI modules require a separate check.

checks

Checks that find no items. Make counts ordinary trigger checks. For n8n, choose the trigger behavior below; productive starts are already counted separately.

An empty Schedule Trigger run counts as an execution. An empty polling-trigger check does not. With no empty checks, both produce the same total. Official counting rules.

Your comparison

Choose a preset or enter your monthly workload.

Monthly-billing estimates from base configurations checked November 1, 2026, not vendor quotes. Each result shows its source date and uncertainty. Historical higher-volume observations are excluded from active selection; larger workloads require a fresh tier check.

Enter your monthly workload to compare the listed plans.

Build your workflow

Estimate your workflow usage

Choose a trigger, add up to eight action steps, and set monthly runs. The comparison uses only monthly base configurations checked November 1, 2026. Larger allowances and required features need a separate quote check. It is an illustrative estimate, not a measured bill.

    Polling vs webhook

    Polling vs. webhooks

    An illustrative two-action workflow handles 500 new leads each month. Scheduled polling checks every five minutes; webhook mode waits for a new lead event.

    The scheduled example has 8,640 total checks, of which 500 find an item. In the calculator, enter only the remaining 8,140 empty checks: productive starts are counted separately. Zapier polling adds no tasks; Make meters ordinary trigger checks and n8n counts scheduled starts.

    Head to head

    The pricing models, side by side.

    Base configurations checked 1 November 2026 with monthly billing selected. September volume observations are preserved as history but no longer power calculations. Make/n8n dollar currency codes remain unverified; annual equivalents are reference only.

    Platform Base plan prices Included allowance One billable unit Polling treatment here Overage approach Self-host Best fit
    Zapier Monthly: Free $0 /mo · Professional $29.99 /mo · Team $103.50 /moAnnual equivalent: Free $0 /mo · Professional $19.99 /mo · Team $69 /mo Free 100 · Professional 750 · Team 2,000 tasks One ordinary action task per completed action. Trigger charges are unresolved and excluded here. No separate polling line; the estimate counts post-trigger tasks. Overage charges are not modeled. Confirm the current rate and account limits with Zapier. No self-host option included here. Fast setup for common app-to-app workflows.
    Make Monthly: Free $0 /mo · Core $16 /mo · Pro $28 /mo · Teams $51 /moAnnual equivalent: Free $0 /mo · Core $12 /mo · Pro $21 /mo · Teams $38 /mo Free 1,000 credits; Core, Pro, and Teams figures shown at 10,000 credits Each action and polling check counts as one credit in this model. Polling checks are added to actions as credits. Not modeled; the estimate stops at the listed plan limit. No self-host option included here. Visual control over branching and data transformation.
    n8n Monthly: Starter $24 /mo · Pro $60 /mo · currency code unverified Starter 2,500 · Pro 10,000 executions; self-hosted Business excluded One whole workflow execution, regardless of how many steps run inside it. Successful starts plus extra empty checks count; batch size can reduce successful starts. Not modeled; the estimate stops at the listed plan limit. Yes. Infrastructure, maintenance, and operating costs are excluded. Technical teams needing flexible logic and deployment control.

    Break-even view

    Ink illustration of two cost curves crossing, one black and one crimson.

    Where each tool wins.

    Compare modeled usage for five ordinary actions per item with no empty checks. The explorer uses the same dated prices as the calculator. A cheapest-tool ranking is withheld when currency or billing rules are unresolved.

    runs

    Illustrative arithmetic: five billed action steps plus the Make trigger per event, zero empty checks, monthly billing; dated source observations below.

    At 1,000 runs, Compare modeled usage.

      Use the explorer above to compare the recorded tiers at your own volume. We do not assume one tool stays cheapest throughout a fixed range.

      Normalization assumptions

      • Five billed action steps per workflow run. So 1,000 single-item runs become 5,000 action tasks in the Zapier model, 6,000 Make credits including triggers, and 1,000 n8n executions. Zapier trigger charges remain unresolved.
      • Zero polling checks and no differently weighted operations. Adding either can move Make’s break-even point earlier.
      • Listed plan limits only. The comparison stops when a plan’s stated task, credit, or execution allowance is exhausted; it does not model overages or larger usage packs.
      • Dated monthly-billing prices. Only the base configurations checked 1 November are selectable; higher-volume prices are not extrapolated. Make and n8n currency codes are unverified, so we do not convert or rank them against USD. Taxes, add-ons, and self-hosting costs are excluded.

      Tracking since September 2026

      Price changes we’ve tracked.

      Dated price observations, not live quotes. Source dates appear with each result.

      Each entry records the observation date and source page. No like-for-like vendor price change was confirmed between the first two snapshots; source corrections and regional display differences are labeled rather than presented as price moves.

      Baseline opened

      • Zapier. Free, Professional, and Team prices were recorded from the official pricing page, with monthly figures and base task tiers corroborated by a secondary source. Baseline
      • Make. Free and Core were recorded from Make’s pricing page; Pro and Teams annual figures came from a secondary source and were marked unverified. Baseline
      • n8n. Starter, Pro, and Business figures were recorded in EUR from secondary sources because the official page did not expose price numbers in that observation. Baseline

      Official-page verification

      • Zapier. Professional $29.99 monthly / $19.99 annual-equivalent and Team $103.50 / $69 were confirmed on Zapier’s official page, with 750- and 2,000-task base tiers. Confirmed no change
      • Make. The official page showed Core $16 / $12, Pro $28 / $21, and Teams $51 / $38 at 10,000 credits. These replaced stale or ambiguous 21 September source entries; the difference is a verification correction, not a confirmed vendor price change. Source corrected
      • n8n. The historical snapshot displayed dollar amounts (the ISO currency was not established): Starter $24 / $20, Pro $60 / $50, and Business $960 / $800. Because the prior snapshot used secondary-source EUR figures, the currency difference is logged as a regional-display caveat, not a confirmed price change. Display changed

      Pricing-page review

      • Zapier. Pricing page restructured around selectable task tiers (100 / 2K / 10K / 100K / 1M / 2M); Professional shown starting from $19.99/month, Team from $69/month; pay-per-task billing now described on-page (exact per-task rate not listed).
      • Make. Core headline displayed $12/mo for 10,000 credits. The earlier $16 monthly and $12 annual-equivalent observations use different billing cadences, so this does not establish a price reduction; page adds AI lineup (Maia by Make, Make AI Agents beta), Make Code App billed at 2 credits per second, and credits overage protection.
      • n8n. Page now highlights AI credits (Starter 2,300/mo, Pro up to 13,700/mo); plan prices not verified on 2026-09-23, no change recorded.

      Trial worksheet

      How to measure your own workflow

      Use these eight cases to check a trial account. Record the usage counter before and after each test, the plan, the date, and any retries. These are suggested tests, not StackReceipt benchmark results.

      1. 01
        Webhook-to-actions. One incoming event followed by a fixed chain of actions.
      2. 02
        Polling. A scheduled check for new records, including empty checks.
      3. 03
        200-row loop. One run processing a controlled batch of 200 rows.
      4. 04
        AI extraction. A repeatable document-to-structured-data workflow.
      5. 05
        Error and retry. A controlled failure followed by the platform’s normal retry path.
      6. 06
        Lead routing. A fixed set of rules directing records to defined destinations.
      7. 07
        Daily digest. Scheduled collection and one consolidated output.
      8. 08
        Filter-heavy. A multi-step flow with several conditions and discarded branches.

      Keep the logic and input volume the same across tools. Subtract the starting usage counter from the ending counter; divide by completed runs only after accounting for failed runs and retries. Record differences in behavior rather than assuming the cheapest counter increase is an equivalent result.

      Migration check

      Should you switch?

      Start with the unit economics, then price the migration itself. These are illustrative differences from the five-action, zero-polling break-even model—not measured bills or guaranteed savings.

      Before switching away from Zapier

      Check a higher Professional task tier before assuming you need Team. Team features and task volume are separate buying decisions. Only base configurations are checked here. A displayed Team allowance is not a claim that Team is required: Professional also offers higher task tiers. Obtain the current Professional quote before choosing.

      When comparing Make and n8n

      Run the same batch in both. Make’s downstream modules can process each item separately; n8n counts starts, including scheduled checks that find nothing. Compare recovery work and required permissions alongside the subscription.

      Before deciding: verify current list prices, rebuild one representative workflow, and compare actual billable units. Annual discounts, taxes, add-ons, overages, engineering time, and downtime risk can outweigh the base-fee difference.

      Migration checklist

      Plan the switch before rebuilding.

      Choose a source and destination to generate a practical sequence. The checklist stays deliberately platform-neutral: it does not assume migration tools or one-click import features.

        Automation guide

        Which tool fits your team?

        Each platform can handle everyday automation. The more useful distinction is how you prefer to build, manage, and pay for the work.

        Zapier →

        Start here for straightforward app handoffs.

        A common use

        A form submission creates a CRM contact, then sends a team message. Start by checking that your exact trigger and actions are supported—not just that the app has an integration.

        What to watch for

        A step on the canvas is not always a billable task. Ordinary successful actions consume tasks. Current official pages disagree about trigger charges; this calculator retains the action-only assumption. Filters and Paths are listed at zero tasks.

        Understand Zapier’s billing and test it with your workflow

        Count completed actions

        In a simple example, 500 leads × 2 successful ordinary app actions = 1,000 action tasks under this model, before any disputed trigger charge. A Filter that stops a lead prevents later actions from running. Don’t count that lead as if every branch completed.

        Replaying an entire run can charge again for actions that previously succeeded. Some AI and Code steps have different task rates; this calculator does not model those exceptions.

        Try before committing

        • Test a valid lead, a duplicate, and an incomplete record.
        • Check the task history against your own count.
        • Confirm the required polling interval, premium apps, and team controls on your chosen plan.
        • Decide who fixes expired connections and failed handoffs.

        Billing source: Zapier task accounting. Fit and trial checklist: StackReceipt editorial guidance.

        Before you buy: timing, retries, and shared connections

        Check the trigger, then the plan

        A multi-step Zap needs a paid plan even at low volume. Filters and searches can make a Zap multi-step despite adding no billable tasks. Premium apps may also require paid access.

        For polling triggers, both the app and your plan constrain the interval. Submit a new record after publishing and measure its arrival; an editor test does not establish live delivery speed.

        Hand over the connections, too

        On Team and Enterprise, the Zap owner needs access to every required app connection. Revoking that access can pause the workflow. Include connections in your handover checklist.

        When recovering a failure, distinguish replaying failed steps from replaying the whole run. Full replay repeats earlier successful actions, so inspect destination records before creating them again.

        Sources: Plan requirements · Trigger timing · Connection ownership · Replay behavior. Trial suggestions are editorial guidance.

        Read the Zapier guide: uses, pros, cons and buyer questions →

        Make →

        Consider it when records need routing and transformation.

        A common use

        New orders are split by region, line items are reformatted, and each order is sent to the right fulfillment team. Map how many records travel down each branch before estimating usage.

        What to watch for

        One scenario run can process many bundles of data. Downstream modules may run once for each bundle, so a short-looking workflow can perform a lot of operations.

        Understand Make’s bundles, operations, and credits

        Follow the records

        For example, one trigger retrieves 10 records, and two downstream modules process every record. That is 1 + 10 + 10 = 21 operations, not three. Operations and credits are different measurements; check the actual credit rate of each module.

        AI features can use different credit rates. The calculator adds a trigger for each successful start and each extra empty check, plus actions across all items. Use batch size when one start processes several items. It does not infer branches, retries, aggregation, or special module rates.

        Try before committing

        • Run a batch, not just one sample record.
        • Inspect each module’s operation and credit counts.
        • Test an empty polling check and an error route.
        • Check whether records need individual actions or can be combined into a summary.

        Sources: Make operations and bundles · Credit accounting. Trial checklist: StackReceipt editorial guidance.

        Before you buy: exact triggers, recovery, and team access

        “Supports Sheets” is not enough

        Make’s Google Sheets Watch Changes module responds to user edits, not API or script updates, and does not watch newly added rows. Watch New Rows is a different module; blank rows can prevent subsequent rows from being processed.

        Test how your data actually arrives before choosing a trigger. A successful manual edit does not prove that an API-written update will start the same workflow.

        Choose how failures are handled

        Incomplete-execution storage is off by default. Enable it deliberately if you need that recovery path, and assign someone to review failures. Automatic retries depend on supported errors or an appropriate handler.

        Connections belong to a team. Plans below Teams provide one team without team-role management. Check who can edit scenarios and inspect their data before sharing access.

        Sources: Google Sheets module limits · Incomplete executions · Teams and roles. Trial suggestions are editorial guidance.

        Read the Make guide: uses, pros, cons and buyer questions →

        n8n →

        Consider it when you need custom logic and a technical owner.

        A common use

        A webhook starts a workflow that calls an internal API, transforms a response, and updates several systems. Decide who will debug the workflow and whether managed or self-hosted deployment fits your team.

        What to watch for

        Execution-based pricing counts workflow runs rather than each step. That does not make hosting, third-party APIs, or operational work free.

        Understand n8n’s execution model and ownership costs

        Separate usage from operations

        For a single workflow with no additional runs, 500 starts mean 500 executions, whether it contains two steps or ten. But a Schedule Trigger running every five minutes starts 8,640 times in a 30-day month—even if most checks find nothing. Enter those starts in monthly runs.

        Starter and Pro are managed options; Business is self-hosted. Monthly capacity and simultaneous capacity are different: the pricing page lists five concurrent executions for Starter and twenty for Pro. If you self-host, assign updates, backups, monitoring, and recovery to an owner. That work is outside this calculator.

        Try before committing

        • Test your hardest API connection and authentication method.
        • Send a burst of requests, not only one at a time.
        • Check what happens when a downstream service fails.
        • For self-hosting, rehearse restoring a backup and name an operational owner.

        Source: n8n execution model and deployment options. Operational checklist: StackReceipt editorial guidance.

        Read the n8n guide: uses, pros, cons and buyer questions →

        LLM-native alternatives

        AI scheduling alternatives

        For recurring work that is mostly reasoning, writing, research, or code, an AI subscription may replace a small automation stack. These tools are complements—not full substitutes for deterministic, multi-app orchestration.

        Claude Code Routines

        Research preview · paid Claude plans · announced Apr 2026

        Best for

        Cloud-run coding and repository work that should recur or respond to an API call or GitHub event—such as maintenance, audits, issue triage, and report generation.

        Where it can win

        When the job is chiefly a prompt plus code context, it can avoid translating the work into a long chain of nodes, app actions, and usage units.

        Limits to weigh

        It is a research preview. Agent output still needs review, and it offers less conventional multi-app branching, connector breadth, and deterministic control than a workflow platform.

        ChatGPT scheduled tasks

        Paid plans · runs while you are away · limit reported around 10 active tasks

        Best for

        Personal recurring prompts: briefings, reminders, summaries, planning check-ins, and other simple work that does not need a visual workflow or many app handoffs.

        Where it can win

        It is quick to set up when the workflow is the instruction itself, especially for someone already paying for ChatGPT. Tasks can run without the app being open.

        Limits to weigh

        The active-task cap is commonly described as about 10, but plan availability and limits can change. Triggers and branching are narrower, and agent reliability varies by request.

        Gemini scheduled actions

        Google AI Pro · limit reported around 10 scheduled actions

        Best for

        Simple recurring research, summaries, and planning prompts for people already using Gemini and paying for Google AI Pro.

        Where it can win

        It removes the cost and setup of a separate workflow service when one schedule and one prompt describe the entire job.

        Limits to weigh

        The scheduled-action cap is reported at roughly 10 and may vary. It is not designed for deep multi-app branching, broad event triggers, or outputs that must be perfectly repeatable.

        Grok Automations

        Evolving from Grok Tasks · introduced early 2026 · access and naming may change

        Best for

        X-native monitoring, trend tracking, and recurring social-listening prompts that benefit from Grok’s access to live X data. Hourly, daily, weekly, or custom schedules can deliver results in-app or, on supported paid access, by email.

        Where it can win

        It keeps the prompt, schedule, model choice, and X context in one place. Prebuilt and custom skills can also make reusable instructions or workflows quicker to run than rebuilding them as nodes.

        Limits to weigh

        Its X-centric advantage does not transfer to every workflow. The product is still moving from Tasks toward Automations, and its multi-app connector story is thinner than Zapier, Make, or n8n. Scheduled automations are described as broadly available, while email triggers may require SuperGrok.

        A practical decision rule

        Choose an LLM-native scheduler when the work is prompt-led, lightly triggered, reviewable, and already covered by a subscription. Choose Zapier, Make, or n8n when you need reliable app-to-app handoffs, branching, retries, structured data movement, and an operational trail.

        Pricing FAQ

        Common pricing questions

        Concise answers based on the same listed-price observations used throughout this page. Observation dates differ by product and tier; see the pricing sources below. These are not live quotes.

        What are Zapier’s listed plan prices and limits?
        On monthly billing, Free is $0 /mo for 100 tasks, Professional is $29.99 /mo for 750 tasks, and Team is $103.50 /mo for 2,000 tasks. Annual-billing equivalents are $19.99 /mo for Professional and $69 /mo for Team.
        What are Make’s listed plan prices and limits?
        Free is $0 /mo for 1,000 credits. At 10,000 credits per month, Core is $16 /mo, Pro is $28 /mo, and Teams is $51 /mo; annual-billing equivalents are $12 /mo, $21 /mo, and $38 /mo.
        What are n8n’s listed plan prices and limits?
        The 1 November 2026 browser check displayed Starter at $24/month for 2,500 executions and Pro at $60/month for 10,000, both billed monthly. The dollar currency code is unverified; this is not a universal USD quote. Business is self-hosted and excluded from the calculator.
        What counts as a Zapier task?
        This model counts each action performed after a trigger as one task. Five ordinary actions therefore use five action tasks in this model. Official sources conflict on whether standard triggers also consume tasks; any trigger charge is excluded and the price ranking is withheld.
        What counts as a Make credit?
        This model counts ordinary triggers plus downstream actions across all items. Batch size groups items under one successful start; extra empty checks add trigger operations. Different credit weights, retries, branches and aggregation are not inferred.
        What counts as an n8n execution?
        One whole workflow run counts as one execution in this model, regardless of how many steps run inside it. That is why n8n can compare differently from per-action pricing on long workflows.
        How does polling affect the comparison?
        Extra empty checks add ordinary Make trigger operations and n8n executions. Zapier polling itself adds no tasks. Enter only empty checks here: successful starts are already derived from items and batch size.
        How are overages handled?
        They are not priced in this guide. The estimate stops at each listed plan limit and flags overage risk. Zapier’s pay-per-task overage rate is not publicly listed, so the page does not invent one.
        Can n8n be self-hosted?
        Yes. Self-hosting can change the economics, but infrastructure, maintenance, upgrades, security, and team time are outside this listed-price comparison.
        Which platform is cheapest at small and large scale?
        In the five-action, zero-polling model, Zapier Free is excluded because multi-step workflows require a paid plan. Use the break-even explorer for the current recorded tiers. Each five-action event contributes six ordinary Make operations, including its trigger. Features and credit weights can change the result.
        How is break-even calculated?
        The explorer models five ordinary actions per event: five Zapier tasks, six Make credits including the trigger, and one n8n execution. It assumes no empty checks, no batching, and monthly billing. It selects among checked base configurations only. Higher Professional task tiers and other unverified configurations may offer different fees; no universal best-fit or cheapest-plan claim is made.
        Should I use the annual prices instead?
        Annual equivalents can lower the recurring fee, but the calculator intentionally uses monthly billing so the comparison does not assume a longer commitment. Verify contract terms and current list prices before deciding.

        Learn

        Automation basics.

        A plain-language field guide to the terms you will meet while comparing, building, and paying for automations.

        API

        An API is a defined way for one piece of software to ask another piece of software for data or an action.

        When Zapier, Make, or n8n creates a contact in another app, it is often making an API request behind the scenes. The API determines what the platform can read or change. A missing API feature can limit an automation even when a platform supports the app.

        Endpoint

        An endpoint is a specific API address for one kind of request, such as finding an order or creating a customer.

        Think of an API as a building and an endpoint as the door for a particular service. Built-in app connections usually hide these addresses from you. In a custom request step, you may choose the endpoint and the data sent to it yourself.

        Webhook

        A webhook is an automatic message one system sends to another when something happens.

        An order system can notify Zapier, Make, or n8n the moment a purchase is completed, so the workflow does not have to keep checking. Webhooks are usually faster and waste fewer usage units than repeated scheduled checks. They work only when the sending system offers the event you need.

        Polling

        Polling means checking an app on a schedule to see whether anything has changed.

        It is the fallback when an app cannot send an instant webhook. A five-minute schedule checks more often than an hourly one, which can affect both speed and usage. This calculator treats polling as a separate input because its billing effect differs by platform; see the billing glossary for the model used here.

        MCP Model Context Protocol

        MCP is a standard that lets an AI assistant connect to tools and data sources through a common interface.

        Instead of building a one-off connection for every assistant and service, an MCP server can describe the tools the assistant may call and the information it may read. That matters here because automation platforms are increasingly becoming the tools AI agents act through—for example, letting an assistant start a workflow that updates a CRM or retrieves an approved record. MCP does not remove the need for permissions, reliable workflows, or human review.

        OAuth

        OAuth is a sign-in and permission process that lets one service access another account without receiving your password.

        It is the familiar “Connect” flow used by many app integrations. You approve a stated set of permissions, and the connection receives a limited token instead of your login credentials. Connections can expire or be revoked, so a working automation may occasionally need to be reconnected.

        API key

        An API key is a secret string an app uses to identify and authorize a connection.

        Some services use a key instead of OAuth when you connect them to Zapier, Make, or n8n. Anyone with the key may be able to use the access it grants, so it should be stored in the platform’s protected credential field rather than pasted into workflow notes. If it is exposed, replace it and update the connection.

        Rate limit

        A rate limit is a cap on how many requests an app will accept within a period of time.

        A workflow can stay within its automation plan and still be slowed or rejected by the connected app’s API limit. Bursts—such as processing a large import all at once—are more likely to hit one. Queues, delays, and careful retries can spread the work without losing records.

        JSON

        JSON is a common text format for organizing data as named fields, lists, and values.

        A webhook might send a customer’s name, order number, and items as JSON. Zapier, Make, and n8n can map those fields into later steps without requiring you to write the format by hand. When a field is missing or nested differently, the workflow may need a parsing or fallback step.

        Integration / app connection

        An integration or app connection is a reusable link between an automation platform and another service.

        It combines authentication with the triggers and actions the platform has prepared for that app. A listed integration does not guarantee every feature of the app is available. Before choosing a platform, check that the exact event and action your workflow needs are supported.

        Trigger

        A trigger is the event or schedule that starts an automation.

        “New form response,” “invoice paid,” and “every weekday at 9:00” are all triggers. A trigger may arrive instantly through a webhook or be found later through polling. The trigger choice affects how quickly the workflow starts and sometimes how usage is counted.

        Action / step

        An action or step is one piece of work performed after the trigger, such as sending a message or updating a record.

        One run can contain several steps, including conditions and data changes. Zapier and Make generally meter work at the task, action, module, operation, or credit level, while n8n’s listed hosted plans in this guide meter complete executions. See the billing glossary for the exact counting assumptions used by this page.

        Scenario / workflow / Zap

        A scenario, workflow, or Zap is the full automation: its trigger, rules, and actions from start to finish.

        Make calls it a scenario, n8n calls it a workflow, and Zapier calls it a Zap. The names differ, but each describes the saved logic that runs when its trigger fires. Compare what happens inside the automation, not just what the product calls it.

        Error handling / retries

        Error handling defines what a workflow should do when a step fails, while a retry attempts that step again.

        Failures can come from expired connections, invalid data, app outages, or rate limits. A reliable workflow records the error, avoids duplicating completed work, and sends truly stuck cases for review. Retry controls differ across Zapier, Make, and n8n, so inspect the behavior you need before moving a critical process.

        Billing glossary

        What the units mean.

        The three platforms price different units. Translating them clearly is the first step toward a fair comparison.

        Zapier task
        One action completed after a trigger. A workflow that performs five actions uses five tasks each time it runs; the trigger itself is not counted in this model.
        Make credit / operation
        One ordinary module operation, including a trigger check, in this model. Some operations may have different credit weights, so the estimate uses the plain one-operation-equals-one-credit case.
        n8n execution
        One complete workflow run, regardless of the number of steps inside it. Five steps in one run still count as one execution here.
        Polling check
        A scheduled check for new data when an instant trigger is unavailable. Make counts ordinary trigger checks. n8n counts empty Schedule Trigger runs, but not empty polling-trigger checks; select the matching behavior in Advanced inputs. Productive starts are already counted from items and batch size. The Zapier help article excludes polls from tasks.
        Overage
        Usage beyond a plan’s included allowance. Overage charges are not modeled; this research has not verified account-specific overage terms.

        Price-change alerts

        Price alerts are coming soon.

        Alerts activate when the notification service launches.

        Reference methodology

        How the estimates work

        • This model treats ordinary successful Zapier actions as one task each. The help article updated 28 October 2026 excludes triggers, polls, Filters and Paths. The rates page still gives conflicting trigger wording (both checked again 1 November), so the assumption stays explicit and we do not rank Zapier by price. Free built-in steps are excluded; special AI and Code rates are not modeled. More than one action per run excludes the Free plan here. Premium apps and other feature requirements can still require a different plan.
        • Make includes triggers and per-item actions. Successful starts = items ÷ batch size, rounded up. Ordinary credits = successful starts + extra empty checks + total downstream actions. Each entered action is assumed to be an ordinary module consuming one credit per input bundle. Filters and routers are free; searches, aggregation, retries, Code and AI require their own accounting. Module rules checked 30 October 2026.
        • n8n counts whole production workflow executions. With a Schedule Trigger, executions = productive starts + empty checks. With a polling trigger, empty checks are excluded. Each batch shares one productive start. Trigger-specific rules checked 30 October 2026. The calculator compares recorded Starter and Pro allowances; Business is self-hosted and excluded. Concurrency and other feature requirements are not modeled.
        • Use-case presets are illustrative starting points, not measured industry data. Each combines active workflows, base monthly runs, average actions per run, and a polling-overhead rate. Company size multiplies only the base runs—Solo 0.35×, Startup 1×, Mid-market 4×, Enterprise 12×—then actions and polling are derived from the scaled runs. Every loaded total remains editable.
        • Calculated amounts are open-ended estimates. The displayed starting amount is the dated fee of a checked base configuration covering the modeled units. It is not a market-wide minimum: unmodeled tiers or discounts may cost less. The + means the upper bound is unknown: feature requirements, taxes, add-ons, and differently weighted operations can increase costs. Another plan’s fee is not used as a ceiling. Only the recorded tiers are compared; other tiers may be available.
        • Listed prices are observations; calculator results are estimates. This is not a live quote. Confirm current prices, currency, billing cadence, and requirements with the vendor.

        What changed, and what remains unresolved

        • 1 Nov 2026Calculator scope corrected: rechecked monthly base configurations, removed historical high-volume tiers from active selection, and changed the headline to workload units rather than a promised price winner. Zapier’s USD selector is explicit; Make/n8n still show only a dollar symbol. The Zapier rates/FAQ contradiction persists. Read the check record. Found a discrepancy? Send the plan, source URL, billing selection, and observation date.
        • 30 Oct 2026Limited refresh, not a price-change claim: browser checks reconfirmed Make 10k monthly tiers and n8n Starter/Pro monthly tiers; currency codes remain unresolved. Usage sources now distinguish n8n polling triggers from schedules and preserve Zapier’s conflicting trigger guidance. Evidence and limits.
        • 23 Sep 2026Volume-tier refresh: Zapier Professional through 200,000 tasks; Make paid tiers through 500,000 credits. Monthly and annual-equivalent observations are recorded separately in the evidence file. Make’s dollar currency code and browser/extracted-page price differences remain unresolved. Make 10k-credit and n8n calculator base tiers were checked again 30 October; higher-volume observations and entries below remain historical.
        • 21 Sep 2026Baseline captured from a mix of official and clearly identified secondary sources. It established the first comparison point; unverified fields were not treated as confirmed vendor changes.
        • 22 Sep 2026Zapier: official pricing page. Recorded monthly and annual-equivalent base prices, 750- and 2,000-task tiers, and the absence of a public pay-per-task overage rate.
        • 22 Sep 2026Make: official pricing page. Recorded monthly and annual-equivalent prices at 10,000 credits, included allowances, and the monthly/annual billing state.
        • 22 Sep 2026n8n: official pricing page. Historical dollar-display snapshot. An earlier USD interpretation was not established; the 30 October browser check still did not establish its dollar currency code.
        • Break-evenFive billed action steps per event plus one ordinary Make trigger, zero empty checks, monthly billing, and the recorded plan allowances. Results are illustrative arithmetic on listed prices, not measured bills.

        Pricing notes

        Pricing sources and observation dates

        Pricing provenance

        1. 01Zapier. Official pricing page, checked 1 November 2026 with Pay monthly and USD selected: Free $0 / 100 tasks, Professional $29.99 / 750 tasks, Team $103.50 / 2,000 tasks. These are base configurations, not every Professional/Team allowance. The rates page includes triggers while the pricing FAQ excludes them; sensitivity cases remain explicit.
        2. 02Make. Official pricing page, checked 1 November 2026 with Monthly billing and 10,000 credits: Core $16, Pro $28, Teams $51; Free $0 / 1,000 credits. Prices were read after the animation settled. Dollar currency code remains unverified. Higher-volume tiers are no longer selected.
        3. 03n8n. Official pricing page, checked 1 November 2026 with Monthly selected: Starter $24 / 2,500 executions, Pro $60 / 10,000. Dollar currency code remains unverified. Business is self-hosted and excluded. Scoped check record and limits.

        Listed prices are dated observations. Calculated results are estimates, not measured bills or vendor quotes. The calculator and break-even view use the monthly-billing figures observed above; annual equivalents are shown only for reference. Taxes, add-ons, self-hosting costs, overages, and usage beyond listed plan limits are excluded. Current vendor pricing may differ from these snapshots.