Contents
At a glance
- Unit
- One run - the unit Stack AI's plan allowances are denominated in
- Free plan
- 500 runs per month at $0 (checked August 2026)
- Enterprise plan
- Custom number of runs, custom pricing
- Published definition of a run
- Not published in prose on the pricing page
- Intermediate tiers
- Not published - the page lists Free and Enterprise
- Model tokens
- Not stated as included - establish separately
What Stack AI counts
Stack AI meters its plans in runs. As listed on Stack AI's pricing page in August 2026, the Free plan includes 500 runs per month at $0, and the Enterprise plan carries a custom number of runs at custom pricing.
The page carries a FAQ question titled "What is a run in StackAI?" but does not render the answer in the page content itself. Stack AI does not publish a definition of a run in prose on its pricing page, and this entry will not invent one. What can be said with confidence is the plan allowance and the fact that runs are the unit those allowances are denominated in.
Why the definition matters more than the number
Five hundred runs a month is either generous or trivial depending entirely on what a run is, and that is the point. If a run is one execution of a whole workflow - the n8n execution model - 500 covers a small production deployment. If a run is one node or one model call - closer to a Zapier task - then a single multi-step agent conversation could consume several, and 500 is a fortnight of light testing.
That is a hundredfold range on the same headline number, which is exactly why the unit definition is the first thing to establish on any platform and why publishing it is a mark of good faith. Ask Stack AI directly before you plan around the allowance.
What to establish before committing
Four questions settle it. Does one run mean one workflow execution regardless of steps, or one step? Do retries and failed runs count? Do runs triggered from the builder while testing count against the allowance, or only production traffic? And are model tokens included in the run, or billed separately by whichever provider the workflow calls?
That last question applies almost everywhere and is the one most often assumed away. On most platforms the platform meter and the model bill are two separate invoices - see token - and a run allowance that says nothing about tokens is almost certainly not including them.
What you should know
The gap between the free tier and Enterprise with nothing published in between is itself information: Stack AI is sales-led above the free tier, so a self-serve cost comparison against Make or n8n is not possible without a conversation. Budget procurement time accordingly, and see proof of concept for how to structure a trial that produces a real usage number to negotiate from.
Stack AI run FAQ
What is a run in Stack AI?
Stack AI meters plans in runs but does not publish a prose definition of one on its pricing page, which lists the question without rendering the answer. Establish directly whether a run is one whole workflow execution or one step before planning around the allowance.
How many runs does the Stack AI free plan include?
500 runs per month at no cost, as listed on the pricing page in August 2026. The Enterprise plan carries a custom number of runs on custom pricing, and no intermediate tiers are published on that page.
Is 500 runs a month enough?
It depends entirely on the unit definition, which is not published. If a run is a whole workflow execution, 500 supports a small production workload. If it is a step or a model call, a handful of multi-step agent conversations could consume it. That hundredfold range is why the definition matters more than the number.
Do model tokens count against Stack AI runs?
Not stated. On most platforms the platform meter and the model provider's token bill are separate invoices, and a run allowance that says nothing about tokens is unlikely to include them. Confirm this specifically before budgeting.
Does Stack AI publish mid-tier pricing?
The pricing page lists a Free plan and an Enterprise plan with custom pricing. Anything between them is a sales conversation, so a self-serve cost comparison against a published-price platform is not possible without contacting them.
Primary source
Checked August 2026. Anything priced or rate-limited moves, so verify against the source before you budget.
Want this built rather than explained?
Knowing what stack ai run means is one thing; having it running in your business is another. We scope one workflow at a time, quote a fixed price in writing after a free consult, build it, and hand it over with documentation you own.
