Amazon Q pricing: every tier, every line item, with the math.

IWhy there are two Amazon Q price lists, not one

The single most common pricing mistake is treating "Amazon Q" as one SKU. It is two products — Q Developer and Q Business — with separate tiers, separate billing mechanics, and separate things you can overspend on. Get the product right first; the numbers follow.

Amazon Q is AWS's brand for generative-AI assistants, and it covers two products that share almost nothing except the name and the security posture. Amazon Q Developer is an AI coding assistant for engineers — it lives in the IDE, the CLI, and the AWS console. Amazon Q Business is an enterprise assistant that answers employee questions over your own company data via retrieval-augmented generation (RAG). Different buyers, different surfaces, and — the reason you are on this page — different pricing. If you are evaluating budget, the first question is always "Developer or Business?" because the price lists do not overlap.

Both products are per-user, per-month subscriptions billed through your AWS account, and both have a low-cost or free entry point so you can pilot before committing seats. Q Developer offers a genuinely free tier for individuals; Q Business starts at a few dollars per user. Beyond the seat, each product has a different "second meter": Q Business charges separately for the index that stores your data, while Q Developer can add usage charges for heavy agentic actions on top of the Pro seat. Those second meters are where real bills diverge from back-of-envelope estimates.

A note on accuracy that applies to this entire page: the figures below are representative as of 2026. AWS adjusts tiers, limits, and per-unit rates periodically, and exact prices vary by Region. Use these numbers to build a model and understand the shape of the bill — then confirm the live rates on the AWS Amazon Q Developer and Amazon Q Business pricing pages before you commit a budget. We flag every place where a number is approximate.

One more clarification that saves confusion: the "Q" capabilities embedded inside other AWS services — Amazon Q in QuickSight, Amazon Q in Connect — are generally bundled into those services' own pricing rather than billed as a standalone Q seat (with the exception that Amazon Q in QuickSight is unlocked by the Q Business Pro tier). This page focuses on the two licensed products you actually buy seats for.

IIAmazon Q Developer pricing — Free vs Pro

Q Developer uses a clean two-tier model: a Free tier for individual developers and a Pro tier at about $19 per user per month for teams. The complexity is not in the tiers — it is in the per-month limits and the usage that can sit on top of a Pro seat.

The Free tier is aimed at individual developers and is genuinely usable for real work, not a crippled trial. You sign in with a free AWS Builder ID — no AWS account required — and get inline code completion across many languages, the chat assistant, and a capped monthly allowance of the higher-cost capabilities: a limited number of agent interactions (the /dev feature-builder and /transform code upgrades), a limited number of security scans, and a cap on AWS-account chat usage. For a solo developer or someone trialing Q, the Free tier often suffices. The constraint is throughput, not capability — hit the monthly caps and you either wait for the reset or move to Pro.

The Pro tier is about $19 per user per month and is the team plan. It raises or removes the Free-tier caps, but its real value is governance and scale: organization-wide license management through IAM Identity Center, centralized policy controls (which features are enabled, IP-based access, reference tracking for code suggestions), higher limits on agent runs and security scanning, and access to the more capable agentic workflows at production volume. If you are rolling Q Developer out to a team, Pro is effectively mandatory because the Free tier is per-individual and has no central administration.

Here is the line item teams miss: certain heavy agentic actions can carry usage-based charges beyond the Pro seat. The clearest example is large-scale code transformation — running /transform to migrate a big codebase across language or framework versions (for instance a Java 8 → 17 upgrade across thousands of files). Each Pro seat includes a monthly allowance of transformation throughput (measured in lines of code processed); exceed that allowance and the overage bills per unit. For most teams writing day-to-day code, the seat fee is the whole bill. For a team doing a one-off mass migration, model the transformation usage separately — it can briefly dwarf the seat cost during the migration month, then drop back to zero.

IIIAmazon Q Business pricing — Lite, Pro, and the index you must not forget

Q Business has two seat tiers — Lite (~$3) and Pro (~$20) per user per month — plus a separate, usage-based charge for the index that stores and serves your data. The index is the single biggest source of "the bill was higher than we modeled," so it gets its own treatment here.

Q Business Lite is about $3 per user per month and covers the core conversational experience: ask natural-language questions, get cited answers grounded in your connected data, basic chat and summarization. It is designed for large populations of light, mostly read-only users — frontline staff, new hires, anyone who occasionally looks something up rather than living in the assistant.

Q Business Pro is about $20 per user per month and is the full feature set: everything in Lite plus plugins and actions (create a Jira ticket, update a Salesforce record, post to Teams — from chat), custom plugins against your own APIs, the Q Apps no-code builder, and access to Amazon Q in QuickSight for natural-language analytics over your BI data. Pro is for power users — analysts, support leads who take actions, sales engineers, app builders.

A genuinely useful cost lever: you can mix tiers inside one application. Put the broad population on Lite and the smaller set of power users on Pro, and the blended per-seat cost drops well below a flat $20. Most well-run Q Business deployments are mostly-Lite with a Pro minority.

IVWhat each tier actually includes — and the upgrade triggers

Pricing decisions are really feature decisions. Below is the honest "what unlocks at each step up" so you assign the cheapest tier that still does the job — and know the specific moment a user needs an upgrade.

The recurring rollout error is binary: either everyone goes on the top tier (overspend) or everyone goes on the bottom tier (power users hit a wall and adoption stalls). The fix is to map each role to the cheapest tier that covers its job, and to know the exact capability that forces an upgrade.

• Q Developer Free → Pro trigger: You move from individuals to a team that needs central license management, policy controls, and reference tracking — or any single developer consistently exhausts the Free monthly caps on agent runs / security scans. The moment you need IAM Identity Center administration, you need Pro.
• Q Business Lite → Pro trigger: The user needs to take actions from chat (plugins), build or use Q Apps, use custom plugins against your APIs, or use Amazon Q in QuickSight for natural-language analytics. If the user only asks questions and reads cited answers, Lite is correct and Pro is waste.
• When the index needs more units: Index capacity, not seats, is what you scale as your corpus grows past ~20,000 documents per unit (or as query throughput demands). Adding connectors and more documents — not adding users — is what drives index cost.
• Q Developer transformation overage: Only relevant during large one-off code migrations that exceed the Pro seat's included transformation allowance (measured in lines of code). For steady-state coding it never fires; for a mass Java/​.NET upgrade month it can be the dominant cost, then disappears.
• What does NOT cost extra: Per-query model inference in Q Business (bundled in the subscription), and normal day-to-day completions/chat in Q Developer Pro. You are not metered per token the way you are with raw Bedrock.
• What quietly costs extra: Q Business index units as documents grow; connector data transfer; the S3/KMS/networking the deployment sits on; and Q Developer transformation overage during migrations. Model these explicitly.

VA worked per-team cost example (both products)

Abstract per-seat numbers are hard to budget against. Here are two fully worked monthly examples — one for Q Developer, one for Q Business — with every line item, using the representative 2026 rates from above.

Example A — Q Developer for a 40-engineer team. Forty engineers, all on the Pro tier for central management and production limits. Seat cost: 40 × ~$19 = ~$760/month. Steady-state coding (completions, chat, normal agent use) is covered by the seat — no overage. In a quarter where they run a one-off Java 8 → 17 migration across a large monolith with /transform, that migration month exceeds the included transformation allowance and adds, say, a few hundred dollars of usage on top for that month only. Steady-state run-rate: ~$760/month; migration-month spike: ~$760 + transformation overage, then back to ~$760.

Example B — Q Business for a 500-person company. A mixed population: 420 light users on Lite and 80 power users (support leads, analysts, sales engineers) on Pro. Seat cost: (420 × ~$3) + (80 × ~$20) = $1,260 + $1,600 = ~$2,860/month for seats. Their corpus is ~600,000 documents across SharePoint, Confluence, Salesforce, Slack, and S3, needing ~30 index units: 30 × ~$100 = ~$3,000/month for the index. Blended total: ~$2,860 + ~$3,000 = ~$5,860/month. Note how the index (~$3,000) is comparable to the entire seat bill (~$2,860) — exactly the dynamic teams under-model when they quote "seats only."

Two lessons fall out of Example B. First, tier mix matters enormously: had all 500 users been put on Pro, seats alone would be 500 × ~$20 = $10,000/month instead of ~$2,860 — the mostly-Lite split saves ~$7,000/month. Second, the index is a real budget line at enterprise document scale; size it to actual document count and grow units as ingestion grows, rather than guessing high on day one.

VIHow Amazon Q's price compares — Copilot, ChatGPT Enterprise, M365 Copilot

Because Q is two products, it competes on price in two different markets. Here is the head-to-head on dollars — with the caveat that price is rarely the only axis, and the cheaper sticker can hide a second meter (like Q Business's index or Copilot's GitHub dependency).

Coding — Q Developer (~$19) vs GitHub Copilot. GitHub Copilot Business is around $19 per user per month and Copilot Enterprise around $39 per user per month; there is also a lower-cost individual Pro plan (~$10/mo). So Q Developer Pro lands essentially level with Copilot Business and well below Copilot Enterprise. On pure seat price they are close at the Business tier — the differentiators are elsewhere (Q's AWS awareness and managed code transformations vs Copilot's GitHub-native ecosystem and broad model choice). Watch the second meter: Q Developer can add transformation usage during big migrations, while Copilot's value is tied to your codebase living in GitHub.

Enterprise knowledge — Q Business (~$3 / ~$20) vs Microsoft 365 Copilot and ChatGPT Enterprise. Microsoft 365 Copilot is about $30 per user per month, on top of an existing Microsoft 365 license — so the true cost is $30 plus the underlying M365 seat. ChatGPT Enterprise is custom-quoted and typically lands in the $40–$60+ per user per month range with seat minimums. Q Business Pro (~$20) undercuts both on the seat, and Q Business Lite (~$3) is dramatically cheaper for light users — but remember Q Business adds the index line, whereas M365 Copilot folds its grounding into the per-seat price (and assumes you already pay for M365).

The honest read on price. For a broad population of light knowledge users, Q Business Lite at ~$3 is the cheapest credible enterprise-assistant seat in this set by a wide margin. For power users, Q Business Pro (~$20) is still below M365 Copilot (~$30) and ChatGPT Enterprise ($40–$60+). On coding, Q Developer is at parity with Copilot Business and cheaper than Copilot Enterprise. But "cheapest sticker" is not "cheapest in production": include Q Business's index, Copilot's GitHub dependency, and M365 Copilot's underlying-license assumption before declaring a winner. Q's consistent structural advantage is that it bills and runs inside your AWS account, so AWS credits apply to it.

VIIHow to optimize and forecast your Amazon Q bill

Once the tiers are clear, most of the savings come from a short list of deliberate choices. These are the levers that actually move the bill — and the traps that inflate it.

• Mix tiers, default to the cheaper one: For Q Business, default the company to Lite and upgrade individuals to Pro only when they need actions, Q Apps, or QuickSight. This is the single biggest lever — it can cut the seat bill by more than half versus a flat Pro rollout.
• Right-size the index, then grow it: Provision index units to your actual document count (~20K docs/unit) instead of guessing high. Add units as ingestion grows. Avoid indexing low-value or duplicate content that inflates unit count without improving answers.
• Isolate transformation spikes: For Q Developer, treat large code-migration months as a separate, temporary cost line. The transformation overage spikes during the migration and returns to zero — budget it as a project cost, not a run-rate.
• Pilot before buying seats: Use Q Developer's Free tier and a small Q Business pilot (few seats + one index unit) to validate value and measure real usage before committing org-wide seat counts. This avoids paying for adoption you haven't proven.
• Forecast seats from real adoption: Buy seats against measured active users, not headcount. Many deployments over-provision Pro seats for users who turn out to be light/Lite users. Reconcile monthly.
• Let AWS credits absorb the pilot: Because Q bills inside your AWS account, AWS credits apply to it. Founders can apply for Activate credits themselves. For a build, a vetted AWS partner can file POC or incremental workload (IW) funding when the work qualifies and AWS approves.

VIIIHow AWS credits and partner funding change the cost

The prices above are list prices. Amazon Q bills through your AWS account, so AWS credits apply to it. Two kinds of AWS money can help, and different people apply for each.

Partner-filed funding (the bigger route for most startups). A vetted AWS consulting partner can file proof of concept (POC) funding for a pilot, or incremental workload (IW) funding if you already run on AWS and are launching new work. For startups that is often $50K to $150K. Depends on the workload; AWS approves. The partner files it, not you. See AWS POC funding explained.

AWS Activate credits (you apply yourself). Activate Founders is $1,000 to start, up to $5,000, and you apply on the AWS site. Activate Portfolio is up to $200,000 if a VC, accelerator or angel group that is an Activate Provider gives you its Org ID. A consulting partner cannot file Activate for you.

CloudRoute matches you with one vetted partner and is free for startups. Partners pay CloudRoute a monthly membership plus 10% of AWS-approved funding. The list prices on this page are your run rate once credits run out, which is what the tier-mix and index-sizing levers above are for.