The new Zapier
Build. Run. Improve. Repeat.
Build however you want
Build Zaps in Zapier's editor; with code; or by asking Claude, ChatGPT, or Cursor. However you start, with Zapier you can connect to the apps your business runs on.
Keep your workflows working
Automate work when you close your laptop. Workflow Manager shows every step and run, while retries and run history make issues easier to understand and resolve.
Automate more reliably and efficiently over time
Zapier identifies which steps really need AI (and which don't) so over time your workflows run more reliably and affordably.
Let Zapier patch issues automatically
If something breaks (because software does sometimes), Zapier can automatically catch the problem, fix it, and keep your workflow running.
Adjust automatically
The Fluid Determinism difference
Fluid Determinism is Zapier’s unique approach to cost-effective, reliable AI. You build with natural language; Zapier suggests where to use AI and then adjusts your workflow as it runs—reducing token use for steps that run the same way every time but adding it back when the unexpected happens.
Deterministic:
Ticket intake
You set the steps. Every ticket gets an auto-reply and goes to a rep, the same way every time.
See what builders are doing with Zapier MCP
Implementing AI in customer service is not about replacing human agents. It’s about giving customers the right answer as quickly as possible [freeing] up our human agents to focus more on our customers with complex cases.
47,000
tickets resolved entirely by Zapier each month
3,000+
agent hours freed each month
Transform dates, currency, text, and more into your preferred format.
Add advanced logic, code editing, and workflows that can improve themselves over time with Next Gen Zaps — available now in early access.
Next Gen Zaps are free during early access, and task usage won't count against your limits. Start exploring whenever you're ready.
Live now
In early access
FAQs
Nothing beyond your existing plan. While Next Gen Zaps are in Early Access, building and running them is free on any paid Zapier plan. No task usage, no billing, no surprises. We want you to build real workflows and help us make this perfect.
Next Gen Zaps use the same task-based pricing as the rest of Zapier, and the published price list is unchanged. Standard app actions, like sending a Slack message or updating a Salesforce record, cost one task each, the same as they cost in a Zap today. The one new line item is the fetch call your agent uses to read data from an app, at one task per call. You only pay for actions that successfully run. Agentic building, editing, execution and management are also free during this period.
Rates are not changing. An action that costs one task in your Zap today costs one task in a Next Gen Zap. Bigger workflows use more tasks because they do more work, so a Zap that loops over thousands of records will use more tasks than a two-step Zap. What changes is what you can build: loops over thousands of records, branches that split and rejoin, custom code steps, and error handling that keeps going.
No. Composing, editing, and testing a Next Gen Zap is free, whether you build it with your agent, in the visual editor, or in code. Failed runs, filter runs, and halted runs are free, too. You pay for successful runs only, during Early Access and after.
Next Gen Zaps can include custom code that calls the apps and APIs you need. Each call your code makes to an outside service counts as one task. Logic that runs inside your workflow, like filters, formatting, loops, and branching, is free.
Not before we tell you. We will announce pricing details and a clear start date while we refine Next Gen Zaps with the Early Access community. There are no penalties for learning: nothing you build or run during Early Access will generate a bill.
