The question people ask before they automate anything isn't "can it do this?" It's "what will this cost once it's actually working?" With the biggest automation tools, the honest answer is "it depends on how many steps you build," which is a strange thing to be charged for. Add a check to catch a bad invoice and the bill goes up. Add a log line so you can see what happened and it goes up again. Invite the colleague who should review the workflow and you're buying another seat.
Lodol is priced the other way around. You pay for runs, and a run is the same price however carefully it was built.
Starting October 1, every plan includes a monthly allowance of runs and unlimited seats. Builder will start at $18 a month and Pro at $120, and more workflows will be able to run at the same time on every plan: 3 on Free, 10 on Builder and 40 on Pro. If you already subscribe, you keep the price you pay today, and your plan will include its standard allowance: 5,000 runs a month on Builder and 10,000 on Pro. You can choose a different amount at any time.
What counts as a run
A run is one pass of a workflow from start to finish, however many steps it takes. A five-step workflow and a fifty-step workflow each count once. Runs you start count, test runs from the editor included, and so do runs started by a schedule, a trigger such as a new email, or the Developer API.
A few things don't count. A workflow sitting and waiting for its trigger costs nothing until the trigger fires. And when one workflow calls another, only the first counts, so splitting a big process into smaller reusable pieces never doubles the bill.
The plans
Every plan has unlimited seats, so you can invite the whole team, reviewers and approvers included, without counting heads.
| Plan | Price a month | Runs a month | Running at once |
|---|---|---|---|
| Free | $0 | 200 | 3 |
| Builder | From $18 | 2,500 to 50,000 | 10 |
| Pro | From $120 | 10,000 to 250,000 | 40 |
| Enterprise | Agreed with us | Custom | Custom |
Free comes with community support. Builder adds email support and higher API limits, and on every paid plan, a trigger that watches an app, such as a new email, checks every minute rather than every 15 minutes. Pro adds advanced permissions and an activity audit log you can export. Enterprise adds custom API limits, SLAs and dedicated onboarding.
On Builder and Pro you choose how many runs you need and the price follows. Builder is $18 a month for 2,500 runs, $30 for 5,000, $54 for 10,000, $120 for 25,000 and $210 for 50,000. Paying annually gives you two months free on every plan, so Builder with 5,000 runs is $300 a year instead of $360. Free needs no credit card, and on every plan your workflows stay private to your workspace until you share one.
What happens if you run out
Nothing surprising. Runs reset on the 1st of every month (UTC), and there are no overage charges, ever. The workspace owner gets an email when 80 percent of the month's runs are used. If a workspace uses them all, workflows stop starting new runs until the 1st, and each paused workflow says so in your Runs list rather than going quiet. Events that arrive in the meantime, like new emails, aren't saved up to run later, which is what the 80 percent email is for: add runs then, from Billing, and nothing is missed. More runs apply straight away, prorated for the rest of your billing period.
How to work out what you need
The number that matters most is how often each workflow runs. On a 30-day month:
| A workflow that runs | Runs a month | Smallest plan that covers it |
|---|---|---|
| Once a day | 30 | Free, or Builder at $18 |
| Every hour | 720 | Builder at $18 |
| Every 15 minutes | 2,880 | Builder at $30 |
| Every 2 minutes | 21,600 | Builder at $120 |
Workflows that start on a trigger run once per event. One that wakes for every new email in an inbox runs once for every email that arrives, including the ones it then ignores, so naming the sender or domain on the trigger is the easiest way to keep the count down. A trigger usually costs far less than a schedule that checks every few minutes, because it only runs when something actually happens.
Put together, a typical small team's set might look like this: a morning check of the inventory sheet (30 runs), an alert in Slack for every email to a shared inbox that gets about 40 a day (1,200), and an hourly sweep of the orders inbox (720). That's 1,950 runs a month, which fits in Builder's $18 tier with room to spare.
What about AI?
A workflow only uses a model at the steps that ask for one, and neither kind of AI step adds a Lodol charge. A judgment call, a condition written in plain words such as "is this email a sales lead?", runs on Lodol's account and is included in the run. An AI step that reads or writes something, such as a receipt or a reply draft, runs on your own Anthropic, OpenAI or Gemini account, and your provider bills you at its own rates. Lodol adds nothing on top.
Compare that with handing the whole task to an AI agent every time. The agent pays for tokens on every run, working out the same steps again and again. Built once as a workflow, every run after that follows the same steps, and a model is paid only where judgment is actually needed. The calculator on our homepage puts numbers on the difference for your own tasks.
Why we price it this way
Per-step pricing quietly punishes the habits that make automation trustworthy. The check that catches a duplicate payment, the step that waits for a manager's approval, the line that records what happened: each one costs more on a meter that counts steps. We want people to build their workflows carefully, so care is free.
Seats are the same story. A workflow that pays vendors should be reviewed by the person who owns that decision, and they shouldn't need a license to do it. Automation works best as something a whole team can see into, and a price per seat works against that.
Start free, upgrade when it earns it.
The Free plan includes 200 runs a month and needs no credit card. See every plan and run tier on the pricing page, or create your workspace.
