In this article
Start with Zapier’s own usage and pricing controls
For an existing account, open Billing and usage and check total billable tasks, your task tier, billing interval and pay-per-task settings. For a new workflow, estimate its successful actions, then select that volume on the official Zapier pricing page.
These are public USD starting prices checked on September 11, 2026. A starting price is not a quote for every task volume. Move the task selector before budgeting for production.
| Platform plan | USD starting price |
|---|---|
| Free | $0; 100 tasks/month and two-step Zaps |
| Professional | $29.99 monthly or $19.99/month billed annually |
| Team | $103.50 monthly or $69/month billed annually |
| Enterprise | Custom quote |
At the displayed entry rates, twelve monthly Professional payments total $359.88, versus $239.88 at the annual equivalent. Team is $1,242 versus $828. These calculations exclude overages, add-ons and applicable tax. They only compare the displayed starting tier.
Professional adds multi-step Zaps, premium apps and webhooks. Team adds shared connections, SAML SSO and access for 25 users. Enterprise pricing depends on negotiated capacity and organization requirements.
A Zap run can consume several tasks
A Zap is an automation workflow. A task is a billed unit of successfully completed work. One new lead can trigger several successful actions, so “500 leads per month” does not automatically mean “500 tasks.”
Zapier’s task measurement guide, updated August 21, 2026, says ordinary successful action steps count toward usage. Triggers, Filters, Paths and actions that fail or halt do not. Replaying an entire run can count successful steps again.
Source: Zapier Help Center, task usage measurement, captured September 11, 2026. The definition and examples show why successful actions, including repeated steps, affect the bill.
For example, a new form submission that creates a CRM contact and sends an email uses two ordinary action tasks if both actions succeed. Five hundred such submissions would use 1,000 tasks. This is a worked estimate, not a measurement of your account; branching, retries and other steps can change it.
AI, code and MCP have their own task rates
The current task usage rate card makes the important exceptions explicit:
| Work performed | Current task rate |
|---|---|
| AI by Zapier: Standard model | 1 per AI step; tool calls unavailable |
| AI by Zapier: Advanced model | 3 per AI step and 3 per tool call |
| AI by Zapier: Premium model | 5 per AI step and 5 per tool call |
| Bring your own AI model | 1 per step and 1 per tool call, plus your provider’s bill |
| Code by Zapier | 1 per step; extra runtime can add tasks |
| Zapier MCP | 2 per successful tool call |
Thus, 500 runs × (one CRM action + one email action + one Advanced AI step) = 2,500 tasks, assuming the AI makes no tool calls. Add one Advanced-model tool call to every run and the estimate becomes 4,000 tasks. A five-step diagram alone cannot tell you which total applies.
For Code, the runtime billing documentation explains the plan’s standard allowance and additional tasks for extended execution. Treat a slower code step as a separate budget variable when increasing volume.
MCP lets an AI client call connected tools without a conventional Zap. Fifty successful MCP calls use 100 tasks at the published rate. Zapier’s official July 23, 2026 tutorial shows the calendar workflow behind those calls. It demonstrates setup and actions; it does not replace the current rate card or quote the AI client’s subscription cost.
Pay-per-task billing has a ceiling
Zapier’s pay-per-task billing guide, updated August 15, says the feature is available on paid plans and enabled by default for accounts created after January 2024. Older accounts may need to opt in.
After the included allowance, additional tasks cost more than the base rate. Check the rate displayed for your plan and billing interval; a universal multiplier copied from an older pricing guide is not a reliable invoice estimate.
The documented ceiling is three times the plan’s task allocation in total. On the guide’s 750-task example, that means 750 included tasks plus up to 1,500 additional tasks, for 2,250 total. It does not mean 750 included plus 2,250 extra.
Professional and Team account owners can disable the feature in billing settings. Turning it off does not erase charges already incurred. Choose the setting deliberately: continued operation and a fixed spending boundary are different priorities.
Find the workflow responsible before upgrading
Zapier’s usage-location guide, updated August 21, distinguishes the account’s billing total from operational reports. Billing and usage provides the account-wide figure. Zap History helps identify expensive workflows, but its task-usage tab does not include MCP.
Use this sequence:
- Read the current billing-period total and allowance.
- Find the Zaps responsible for most usage.
- Check repeated actions, whole-run replays, AI tiers and tool calls.
- Include MCP separately when estimating from workflow reports.
- Compare the next task tier with the overage rate shown in the account.
If work pauses at the limit, Zapier’s limits guide says held runs can be replayed once capacity is available. Review that backlog before assuming a plan change has completed all missed work.
Platform, Agents and Chatbots also appear as separate pricing tabs. Do not equate an Agents activity allowance with the platform task allowance. If your automation also relies on paid collaboration tools, include their separate bills; our Slack pricing and Airtable pricing guides explain those models.
Track a published rate change you care about
Use Zapier’s native usage reports and limit notifications for your own account. PageDog becomes useful when you want to watch a public pricing rule between budget reviews.
For example, open the official rate card in PageDog on a Mac, select the MCP rate and its label, and set: “Alert me if Zapier MCP costs more than 2 tasks per successful tool call.” This is a specific public-page condition with a clear baseline.
PageDog checks the selected visible area on your Mac, evaluates the condition with AI, and keeps before-and-after evidence to review. It does not meter Zapier’s private task consumption, manage Zaps or enforce a spending cap. Checks require PageDog open and the Mac awake and online; see the runtime and AI-data documentation.
Choose a plan for the workflow features you need, then size its task allowance from representative successful runs. Keep native billing controls as the authority for spending. Revisit the estimate when the workflow’s model tier, tool calls or execution volume changes.
Sources checked
Plans: Zapier pricing. Metering: task measurement, rate card and Code runtime. Account controls: pay-per-task billing, usage reports and limits. All checked September 11, 2026.
