Amazon Quick Suite Pricing: Cost per User, License Tiers and the $250 Account Fee

Amazon Quick Suite costs $0, $20 or $40 per user per month depending on plan, plus a flat $250 per account per month on the two business plans. That is the whole price list. Everything else is metered overage on top.

The naming is worth clearing up first, because it is the reason this is hard to search for. AWS has dropped the word "Suite": the product page now sits at aws.amazon.com/quick/, the old aws.amazon.com/quicksuite/ address returns a 301 redirect to it, and the pricing page title reads "Amazon Quick Cost". Most buyers and most of the internet still say Quick Suite. Both names refer to the same product, and every price below was read directly from the AWS pricing page on August 22, 2026.

Look at the Plus and Professional rows next to each other. Same seat price, same $20. The only difference in what you hand AWS is $250 a month, flat, forever, no matter how many people are on the account. That single line decides more small and mid-size Quick deals than any feature on the comparison chart, and it is the first thing worth modelling.

AWS states the fee plainly: "Accounts are charged a flat $250 per account per month infrastructure fee. This fee covers the underlying AI infrastructure that powers the Amazon Quick experience." Flat means flat. It does not scale down for small teams and it does not scale up for large ones, which makes it either trivial or brutal depending entirely on headcount.

Here is what it does to the real cost of a Professional seat once you divide it across the people actually using it.

A ten-person team on Professional is not paying $20 a seat. It is paying $45 a seat, which is more than Microsoft charges for a 365 Copilot licence and more than any reseller-quoted Gemini Enterprise Standard seat. At 125 seats the fee finally falls to 10% of the bill, and past 250 it rounds away to noise. Quick is priced for organizations, and the fee is the mechanism that makes that true.

Enterprise absorbs the fee faster because the seat is twice the price, but the shape is identical.

One practical consequence: consolidate accounts. Because the charge is per AWS account rather than per organization, three departments each running their own Quick account pay $750 a month in infrastructure fees for the same product one consolidated account would cover for $250.

An agent hour is metered compute time for anything AI-driven in Quick. AWS defines it as "time used by AI features in Quick Research, Quick Flows, Quick Automate, the desktop app, artifact generation, and custom AI-powered apps," and says usage is metered per second, so you are billed for what you actually consume rather than in rounded blocks.

The allowance is split into two buckets, and the split matters because the two buckets cost different amounts once you exceed them.

AWS allows some give between the buckets: "users who exceed their agent hour entitlement can consume into their research hour entitlement," so the two allowances behave as one pool with a preferred order rather than as hard separate caps.

The allowance is priced at 90% of the seat

Here is a calculation AWS does not publish but which falls straight out of its own rate card. Value the included hours at the overage rates you would otherwise pay for them:

Both tiers land on exactly 90%. That is not a coincidence, and it tells you how AWS expects the product to be used: the seat fee is close to pure allowance, and any team doing serious automation is expected to buy hours on top. Two useful thresholds follow from it. On a $20 Professional seat, 6.67 hours of flow or automation overage doubles that user's cost, and only 3.33 hours of research overage does the same. Deep research is the expensive verb in this product.

Note also that the entitlement is per subscriber. AWS phrases the allowance as belonging to "Professional subscribers" and "Enterprise subscribers," while index storage is explicitly pooled across the organization. Two different sharing models in one bill, so model them separately.

Doubling the seat price from $20 to $40 buys exactly three additional capabilities, double the agent hours and double the index storage. That is the complete list, taken cell by cell from the AWS feature comparison rather than from marketing copy.

The governance story is the surprise here. Professional already includes RBAC, SSO, admin controls, data sovereignty controls, centralized billing and 24/7 support. Security and compliance are not the upsell. The upsell is authoring: building multi-step automations in Quick Automate, creating rather than consuming Quick Sight content, and running an approval workflow over published assets.

Which means most organizations should buy both

Because the three Enterprise-only capabilities are all authoring functions, the natural shape is a small number of Enterprise seats for the people who build automations and datasets, and Professional seats for everyone who only consumes them. AWS lists "unlimited user scaling" on both tiers and charges the $250 fee once per account, so a mixed account is not penalized.

At 50 users, all-Professional costs $1,250 a month and all-Enterprise costs $2,250. Ten Enterprise builders plus 40 Professional consumers costs $400 + $800 + $250, or $1,450, which is $800 a month less than putting everybody on Enterprise for capabilities most of them would never open.

Free and Plus are the no-AWS-account tiers. You sign up with an email address or a social login, and neither carries the $250 infrastructure fee. AWS positions them for "individuals exploring AI-powered productivity" rather than for organizations.

Plus adds exactly five things over Free.

Free is more capable than most free tiers in this category. It includes Quick Flows automation, connectors to Slack, Microsoft 365, Google and QuickBooks, in-depth research and app building. What it withholds is collaboration and packaging: you cannot save a custom agent for others, share a Space with your team, or push a knowledge base across the org.

The honest read for a small business is that Plus at $20 covers a surprising amount of ground, and the jump to Professional is a governance and analytics purchase, not a capability one. If nobody in your organization needs SSO, role-based access, data sovereignty controls or Quick Sight, Professional's $250 monthly fee is buying you administration you are not going to use.

Amazon Quick is the only major AI workspace in this group that charges a flat platform fee per account on top of seats. That is the structural difference buyers should weigh, more than the seat price itself, because it is the thing that changes the answer at different company sizes.

Read across the rows at two different company sizes and the ranking flips. At 10 seats, Quick Professional's effective $45 per user is more expensive than Microsoft 365 Copilot's flat $30. At 250 seats, Quick Professional's effective $21 undercuts it by nine dollars a head, which on 250 people is $27,000 a year. Nothing about the products changed between those two rows; only the denominator did.

Three caveats on that table, stated plainly. Gemini Enterprise seat prices remain reseller-reported rather than published by Google, so treat the $30 as an indication and get a quote. Glean prices by negotiation, and the FlexCredit consumption sits on top of seats. And the platforms are not feature-identical: Quick is the only one of the group that bundles a full business intelligence product, which is why the Quick Sight add-on rates below matter to the comparison.

If you are weighing these against each other rather than against doing nothing, our Glean pricing breakdown, Copilot Studio pricing and Agentforce pricing pages carry the same cell-by-cell treatment for each vendor, and the AWS and Microsoft equivalents of Google Agentspace maps how the three clouds line up feature by feature.

Business intelligence in Quick is partly bundled and partly sold separately. Viewing Quick Sight dashboards and running scenario analysis are included on both paid business tiers. Three capabilities are priced on top.

That Pixel-Perfect minimum deserves attention. At $1 per report unit with a floor of 500 units a month, the smallest possible commitment is $500 per month. To put that in proportion, $500 a month is more than the seat bill for 20 Professional users or 12 Enterprise users. A finance team that wants formatted recurring reports is signing up for an add-on that can cost more than the entire subscription it attaches to.

AWS also notes a "BI-only pricing option" for customers using Quick exclusively for business intelligence, which is worth asking about if dashboards rather than agents are the actual reason you are buying.

How index storage is measured, and why the number is bigger than you expect

AWS measures index storage as "GB of raw storage, the original file size of documents at their source location (such as SharePoint, S3, or Confluence) before any processing." That is the pre-processing size of everything you connect, not the size of the resulting index, so a connected SharePoint site counts at its full weight.

The allowances pool across the organization: 25 GB per user on Professional, 50 GB on Enterprise. A hundred Professional users therefore share 2,500 GB. Overages are $5 per GB per month, which compounds quickly. Exceeding that 2,500 GB pool by ten percent costs $1,250 a month, or 62.5% of the $2,000 seat bill it sits on top of. Audit what you connect before you connect it.

Yes. AWS offers a 30-day free trial of the Enterprise plan for new customers. During the trial the full subscription fee is waived for up to 25 users, the $250 infrastructure fee is waived for the account, and no credit card is required to begin. Customers get access to all Enterprise features, and at the end of 30 days customers with an AWS account start being billed for an Enterprise plan.

Priced at list, that trial is worth $1,250: 25 Enterprise seats at $40 plus the waived account fee. It is a genuinely generous evaluation window, with one condition to diarize. Billing starts automatically at day 31 for anyone holding an AWS account, so the trial converts by default rather than by decision. Put a reminder at day 25.

Use the window deliberately. The two things worth measuring are the ones you cannot get from any price list: how many agent hours your real workflows consume in a fortnight, and how many gigabytes your connectors actually index. Those two numbers turn the rate card above into a forecast. Everything else you can read.

There is a separate Free plan that runs indefinitely with no AWS account and no time limit, which is the better route if you want to evaluate the assistant itself rather than the governance layer.