Amazon Q cost and Amazon Q pricing: Amazon Q Business pricing, Q Developer rates and Quick Suite fees in 2026

It depends which product you mean. Amazon Q Business is $3 per user per month for Lite and $20 for Pro, plus index capacity billed hourly. Amazon Q Developer is $0 for Free and $19 per user per month for Pro. Amazon Quick, the successor, runs $0, $20, $20 and $40 across its four tiers, with a $250 per account per month infrastructure fee on the top two.

The confusion is AWS-made rather than yours. Three products have shared the Q branding, and search results for Amazon Q pricing return all three interchangeably. The quickest way to tell which one you want: if the job is answering questions across your company content, you want the Business or Quick line. If the job is writing and transforming code, you want Developer.

Whichever it is, budget the meter rather than the seat. Q Business bills index capacity by the hour, Quick bills a fixed account fee plus agent hours, and Developer bills per line of code above the included allowance. None of the three bills you only for seats.

No. AWS closed Amazon Q Business to new customers on July 31, 2026 and states that new feature requests will no longer be considered. Existing customers keep their applications, the APIs stay available, and bug fixes and security updates continue. AWS has published no shutdown or end-of-support date.

This puts existing customers in a reasonably comfortable spot and prospective ones in none at all. If you are already running Q Business, nothing breaks and there is no forced migration date on the calendar, so you can plan the move on your own schedule rather than someone else's. What you are giving up is the roadmap, and over a two or three year horizon a product receiving only bug fixes falls behind on connectors and model quality.

Q Business was not alone. AWS moved a group of first-generation AI services into maintenance mode in the same window, including Amazon Kendra and Amazon Bedrock Agents Classic, and in every case the pattern was the same: closed to new customers, existing workloads supported, no shutdown date announced. That is a gentler retirement than a hard end-of-life, and it is still a retirement.

Because Amazon Q Business bills index capacity by the hour, independently of usage. A starter index unit runs $0.140 an hour and an enterprise unit $0.264, so the meter accrues around $102.20 or $192.72 a month even if no one submits a single query. Seats are billed per user per month on top of that.

This trips up teams who pilot the product, lose interest and leave the application running. There is no idle discount and no automatic scale-to-zero. If a proof of concept is finished, deleting the application is what stops the charge, and detaching users does not.

It also means very small deployments have terrible unit economics. Five people on Pro is $100 in seats against roughly $193 for a single enterprise index, so nearly two thirds of the bill is infrastructure. The ratio only becomes sensible in the other direction: at a hundred Pro users the same index unit is about 9 percent of a roughly $2,193 monthly bill.

Roughly $2,193 a month, or about $26,313 a year. That is 100 Pro seats at $20 for $2,000, plus one enterprise index unit at $0.264 an hour, which is about $192.72 over a 730 hour month. Lite seats instead of Pro would bring the same deployment to about $493 a month.

Two things move that figure, and both are about content rather than headcount. The first is corpus size: one index unit covers about 20,000 documents or 200 MB of extracted text, whichever comes first, so 200,000 documents needs ten units and about $1,927 a month in index alone, which would nearly double the total above. The second is media, billed at $0.003 an image, $0.006 a minute of audio and $0.050 a minute of video.

The useful lesson generalizes past AWS. In this category, cost scales with how much content you index far more reliably than with how many people you hire, and a per-seat quote hides that completely.

It is a fixed $250 per account per month charge that applies to the Amazon Quick Professional and Enterprise tiers, on top of per-user pricing. Free and Plus do not carry it. Because it is fixed rather than per seat, it raises effective cost per user sharply at small headcounts and becomes negligible at large ones.

Here is what it does to Professional at $20 a seat:

A ten person team pays 125 percent above the sticker price. A 250 person company pays 5 percent above it. That is not a criticism of the fee, which reflects real fixed infrastructure, but it does mean the advertised $20 describes the experience of a large customer and misdescribes a small one.

Enterprise behaves the same way from a higher base: ten seats at $40 plus $250 is $650, an effective $65 each, settling toward $42.50 at a hundred users. If you are under about 25 people and do not need the AWS account integration, Plus at $20 with no infrastructure fee is the tier to price first.

No. They are separate products with separate pricing that happen to share a name. Amazon Q Developer is a coding assistant priced at $0 for Free and $19 per user per month for Pro. Amazon Q Business was the enterprise search and knowledge assistant at $3 and $20 per user per month plus index capacity.

The overlap is close to zero in practice. Q Developer works on source code: it answers questions about a repository, generates code and runs transformations, and its meter reflects that, with 50 agentic requests a month on Free and an overage of $0.003 per line of code above 4,000 lines per user per month on Pro. It does not index Slack, SharePoint, Confluence or Google Drive, so it cannot answer the question that sends people looking for enterprise search in the first place.

Worth knowing if you are comparing quotes: Q Developer Pro is still openly available while Q Business is not, so a vendor comparison that lists Amazon at $19 a seat is almost certainly quoting the wrong product for a search evaluation.

Answer length and feature access. Lite at $3 per user per month returns permission-aware answers of roughly one page with single sign-on and file insights. Pro at $20 extends answers to roughly seven pages and adds content creation, Q Apps, QuickSight integration, plugins and image processing. Both draw on the same index, so the underlying content and permissions are identical.

The seven-to-one price gap is really about whether people are reading answers or working inside them. Lite suits a large population of occasional lookups, where someone needs a policy confirmed and leaves. Pro suits the smaller group building on top of results.

Mixing tiers is the sensible default, and it is a pattern worth applying to any vendor that offers it. Most organizations have a small core of daily users and a long tail of monthly ones, and putting the tail on the cheap tier costs almost nothing while putting it on the expensive one is where budgets go.

Index capacity, media processing, consumption above the included allowance, the AWS account and IAM work to run it, and on Amazon Quick the $250 monthly account fee and any agent hours past your allowance. None of these appear when someone quotes Amazon Q at $20 a seat.

Take them in order of how often they surprise people. Index capacity is first and largest, because it is hourly and scales with documents rather than users. Consumption is next: $200 buys 30,000 units a month and each ChatSync API call or embedded prompt consumes two, so roughly 15,000 prompts, or about $0.013 each. Media is metered at $0.003 per image, $0.006 per minute of audio and $0.050 per minute of video, which stays small unless you are indexing recorded meetings, at which point an hour of video costs $3.00.

On the Quick side the agent hours are the line to watch, since Professional includes four a month and Enterprise eight, with overage at $3 an hour and research agent hours at $6. Storage past the pooled allocation is $5 per GB per month. There is a 60 day free trial on Q Business for up to 50 users per application including 1,500 index hours, and a 30 day Enterprise trial on Quick that waives the subscription fee for up to 25 users and the $250 infrastructure fee, both of which are genuinely useful for sizing before you commit.