Contents
Fast answer
Stack AI publishes exactly two things as of 20 August 2026: a free plan at $0 a month with 500 runs, 1 seat and 2 projects, and an Enterprise plan at custom pricing with custom runs, custom seats and unlimited projects. There is no self-serve paid tier of any kind between them and no published rate for the Enterprise tier. That means the entire priced range of the product is a 500-run prototype allowance, and everything beyond it requires a sales conversation with no public anchor. The metered unit is runs rather than steps or seats, so cost scales with usage volume rather than workflow complexity, but without a published per-run rate you cannot model that in advance at any volume. For a business that needs more than a prototype and less than a compliance programme, there is currently no priced path, and that gap is the single most important fact about Stack AI's pricing.
Facts last verified 2026-08
Quick facts
- Free
- $0/month, 500 runs, 1 seat, 2 projects (checked 2026-08-20)
- Enterprise
- Custom, with custom runs and seats and unlimited projects
- Self-serve paid tier
- None published
- Metered unit
- Runs, not steps or seats
- Published per-run rate
- None
- Minimum commitment
- Not published
- Trial terms
- The free tier is the trial
- Domain note
- stack-ai.com now redirects to stackai.com
On this page
What this page covers
A comparison visitor should understand the tradeoff, the best-fit scenario, and the next diagnostic tool to confirm the choice.
- 01What is published
- 02What a run means
- 03How to benchmark a quote
- 04Where the meter bites
- 05Cheapest honest path
- 06When to leave
Why does this matter now?
A free tier followed immediately by an enterprise quote is a specific commercial strategy: it filters out everyone who will not sit through a sales cycle. That is a rational thing for a vendor to do and an expensive thing for a buyer to discover in week three of a build. If you are prototyping, the free tier is genuinely usable and costs nothing. If you are planning production, you are planning around a number nobody will show you until you ask, so ask first and build second.
Internal path
Where to go next from this page
These links are part of the A8gent learning and conversion path. Use them to move from concept, to diagnosis, to workflow build, to course.
What you should be able to do after this
- See exactly what is published
- Understand what a run means here
- Benchmark an Enterprise quote against real alternatives
- Decide before entering a sales cycle
How do you do it, step by step?
1. Use the free tier for what it is
500 runs a month, 1 seat and 2 projects is a prototype allowance. It is enough to prove that a document assistant or internal tool works, and it is not enough to operate one. Treat it as a technical evaluation rather than a pilot, and do not let a working prototype create a commitment you have not priced.
2. Understand what a run is before you estimate volume
The meter is runs, not steps or seats, so a complex workflow and a simple one cost the same per execution. Work out how many executions your process would generate in a month, because that is the number an Enterprise quote will be built around, and being able to state it precisely is the strongest position you can take into that conversation.
3. Get the Enterprise number early, not late
Nothing about the Enterprise tier is published: not a per-run rate, not a seat price, not a minimum commitment, not a contract term. Ask for all four in writing before you build anything on the free tier, because the cost of discovering the number after a working prototype exists is that you have already spent the evaluation.
4. Benchmark against platforms that publish
A quote needs something to be measured against. On published August 2026 pricing, Dify Professional is $590 a year for 5,000 message credits a month, Dify Team is $1,590 a year for 10,000, and n8n Cloud Starter is EUR 20 a month for 2,500 executions with self-hosting free. Those are different products, but they set the floor a quote has to justify clearing.
5. Check the connector question separately
Stack AI is built for LLM applications, document assistants and internal AI tools rather than general integration, and it has far fewer general-purpose connectors than mainstream automation platforms. If your workflow needs to touch a dozen business systems, that is a capability question that no pricing negotiation will solve.
The gap between 500 runs and a sales call
Stack AI publishes exactly two things as of 20 August 2026. A free plan at $0 a month with 500 runs, 1 seat and 2 projects. And an Enterprise plan at custom pricing with custom runs, custom seats and unlimited projects. There is nothing between them: no self-serve paid tier of any kind, and no published per-run rate at any volume. That gap, rather than any feature, is the most important commercial fact about the platform.
Five hundred runs a month, one seat and two projects is a prototype allowance rather than an operating one. It is enough to prove that a document assistant or an internal tool works, and it is not enough to run one. One seat means you cannot even trial it as a team, since a second person needs their own access to evaluate anything. So the free tier answers the capability question well and answers the cost question not at all.
The practical failure mode follows directly. A team builds a prototype on the free tier, it works, the business gets interested, and only then does anyone discover that the next step is a sales conversation with no public anchor. By that point the evaluation has already been spent, the internal expectation has been set, and the negotiating position is weak because the alternative is starting over. The fix is to reverse the order: get the Enterprise number before you build, not after.
The metered unit is worth understanding for that conversation. Cost scales with runs, meaning workflow executions, rather than with steps or seats, so a complex workflow costs the same per execution as a simple one. That is a favourable structure for anyone building sophisticated tools at modest volume. Without a published per-run rate, though, you cannot model it in advance at any volume, which is exactly why your own run estimate is the strongest thing you can bring to a quote.
What to measure an Enterprise quote against
Bring published comparators, because a custom quote assessed against nothing is assessed against nothing. Read on 20 August 2026, Dify publishes Sandbox free with 200 message credits, 1 team member and 5 apps; Professional at $590 per workspace a year with 5,000 monthly message credits, 3 members, 50 apps and 500 knowledge documents; and Team at $1,590 a year with 10,000 credits, 50 members, 200 apps and 1,000 knowledge documents, alongside a Community edition that self-hosts free.
n8n Cloud gives the other useful floor at EUR 20 a month billed annually for 2,500 executions with 5 concurrent, EUR 50 for 10,000 executions with 20 concurrent, and a free self-hosted Community edition. Those are different products with different strengths, and that is fine, because their function here is to establish what published pricing looks like for adjacent work. A quote several multiples above that band needs to be earning the difference in capability you will actually use.
Ask for four numbers in writing: the per-run rate or the included run volume, the seat price, the minimum commitment in both spend and term, and the renewal mechanics. Bring your own monthly run estimate so the conversation is anchored on a figure you calculated rather than one assumed on your behalf. A team that walks in knowing it needs 40,000 runs a month is having a different conversation from one that walks in asking what it costs.
Be clear-eyed about what the Enterprise tier is genuinely for, because it is a real offer rather than a paywall. It carries dedicated infrastructure, on-premises and VPC deployment, solution engineers, SSO, access control and SOC 2, HIPAA and GDPR compliance. If you need those, the sales process is part of the purchase rather than friction in it, and Stack AI is a credible option. If you do not, you are entering an enterprise procurement cycle to buy something a published-price platform sells from a signup form.
The capability question a quote will not fix
Before pricing anything, check scope, because it disqualifies more evaluations than cost does. Stack AI is built for LLM applications, document assistants and internal AI tools, and it carries far fewer general-purpose connectors than mainstream automation platforms. If your workflow needs to touch a dozen business systems, that is a capability gap, and no negotiation on run rates changes it. Test the specific integrations you need on the free tier before the commercial conversation starts.
This is the most common category error with this product: treating it as an automation platform because it automates things. A workflow that reads a document, extracts structured fields and answers questions about them is exactly what it is for. A workflow that reads a form, checks a CRM, updates a billing system, notifies a channel and schedules a follow-up is a general integration problem, and n8n at EUR 20 a month or Zapier from $19.99 a month billed annually are built for that shape.
One administrative detail worth knowing so you check the right page. The domain has moved: stack-ai.com now redirects to stackai.com. If you are verifying figures against an older bookmark or a link in a comparison article, make sure you are reading the current site, because pricing pages are exactly the kind of page that gets cached, quoted and then quietly changed.
The honest summary for a small or mid-sized team is that this is a strong prototyping tool with no published path to production. Use the 500 free runs to answer the capability question quickly and cheaply, which it does well. Then either get the Enterprise number in writing before you build anything real, or move to a platform where the second milestone has a price on it. What you should not do is let a working prototype create a commitment you never priced.
At a glance
Stack AI published pricing, read from stackai.com/pricing on 20 August 2026
| Line item | Published price | Note |
|---|---|---|
| Free | $0/month | 500 runs a month, 1 seat, 2 projects |
| Self-serve paid tier | None published | No plan of any kind between free and Enterprise |
| Enterprise | Custom/month | Custom runs, custom seats, unlimited projects |
| Per-run rate | Not published | You cannot model cost at volume in advance |
| Seat price | Not published | Ask for it separately from run volume |
| Minimum commitment | Not published | Ask for spend and term |
| Benchmark: Dify Professional | $590 per workspace per year | 5,000 monthly message credits, 3 team members |
| Benchmark: Dify Team | $1,590 per workspace per year | 10,000 monthly message credits, 50 team members |
| Benchmark: n8n Cloud Starter | EUR 20/month billed annually | 2,500 executions a month; self-hosting is free |
What mistakes should you avoid?
- Building a production workflow on a 500-run free tier and discovering the price afterwards
- Assuming a self-serve paid plan exists between free and Enterprise
- Entering a sales conversation without knowing your own monthly run volume
- Evaluating an Enterprise quote with no published benchmark to compare against
- Choosing it as a general automation platform when the connector catalogue is much narrower than mainstream tools
- Trying to trial it as a team on a plan that allows one seat and two projects
FAQ
- How much does Stack AI cost?
- The free plan is $0 a month with 500 runs, 1 seat and 2 projects. Enterprise is custom priced with custom runs and seats and unlimited projects. As of 20 August 2026 there is no self-serve paid tier between them and no published Enterprise rate of any kind.
- What is a run?
- One execution of a workflow. Because the meter is runs rather than steps or seats, a complex workflow costs the same per execution as a simple one, so cost scales with usage volume rather than complexity. Without a published per-run rate you cannot turn that into a forecast in advance.
- Is there a paid plan below Enterprise?
- None was published when we checked on 20 August 2026. The pricing page showed a free tier and an Enterprise tier requiring a custom quote, with nothing in between. That is the main practical objection for a small or mid-sized team, because there is no priced upgrade path from a working prototype.
- What should I ask for in an Enterprise quote?
- The per-run rate or the included run volume, the seat price, the minimum commitment in spend and term, and the renewal mechanics. Ask before building, and bring your own monthly run estimate so the conversation is about a number you calculated rather than one they assumed.
- What should I compare it against?
- On published August 2026 pricing, Dify Professional at $590 a year for 5,000 monthly message credits, Dify Team at $1,590 a year for 10,000, and n8n Cloud Starter at EUR 20 a month for 2,500 executions with free self-hosting. Different products, but they establish what published pricing looks like for adjacent work.
Sources & further reading
Done for you
Want this built for your business?
Tell us the workflow. We scope, build, and ship the agent with guardrails and the numbers to prove it worked. The scoping call is free.
Free scoping call. You own the code.
Was this page helpful?

