Amazon Quick

What it is

Amazon Quick is AWS’s AI workspace for employees: one chat surface that searches company knowledge, runs research reports, builds dashboards and automates multi-step work. It launched in October 2025 as Amazon Quick Suite, folding the old QuickSight BI product in as Quick Sight, and dropped “Suite” in 2026 when AWS opened self-serve plans that need no AWS account. Today it bundles Spaces (shared collections of files and connected data), Agents (custom chat agents with their own instructions and sources), Research (long-form cited reports), Quick Sight (dashboards and natural-language BI), Flows (no-code repeatable tasks) and Automate (multi-step process automation), with a desktop app, mobile app, browser extension and Microsoft 365 extension around them.

It is also where AWS is sending everyone who bought Amazon Q Business. AWS’s own documentation now says Q Business “is no longer open to new customers”, calls Quick “the next evolution of Amazon Q Business”, and offers a Bring Your Own Index path that attaches an existing Q Business index to Quick as a knowledge base. AWS does not name the models behind Quick on its product or pricing pages; you choose between Fast, Balanced and Smart modes instead.

Why ops teams end up here

The buying question is usually: we are an AWS shop, so does Amazon Quick replace Copilot, Gemini or Glean plus a BI tool, and what do the tiers cost? Quick’s case rests on the bundle. A RevOps team can ask a Space that holds Salesforce, HubSpot and the pipeline dashboard why Q4 coverage dropped, then turn the answer into a Flow that posts the same breakdown to Slack every Monday. A legal-ops team can point an agent at the contract repository in SharePoint or Box with document-level permissions honoured (AWS lists S3, Confluence Cloud, SharePoint and Google Drive for ACL-aware indexing). A support-ops lead can build the ticket-volume dashboard in Quick Sight and let managers query it in plain language, instead of buying a separate BI seat for each of them.

It connects outward through built-in connectors (the product page shows Salesforce, HubSpot, ServiceNow, Jira, Slack, Notion, Box, Zoom and others) and acts as an MCP client, so you can add tools such as Linear, Intercom or PagerDuty through a four-step wizard.

Pricing reality

Five plans, checked on aws.amazon.com/quick/pricing on 29 October 2026:

• Free — $0, one user, 1 GB of index storage. No AWS account needed.
• Plus — $20 per user/month annual, $25 monthly, up to 300 users, 10 GB per seat. No AWS account needed.
• Max — $100 per user/month annual, $125 monthly, 5x the usage and storage of Plus (launched 3 September 2026).
• Professional — $20 per user/month plus a $250 per-account monthly infrastructure fee, 8 agent hours per user/month. Requires an AWS account. Users can view Quick Sight dashboards but not build them, and overage cannot be bought.
• Enterprise — $40 per user/month plus the same $250 fee, 18 agent hours per user/month, dashboard authoring, multi-step Automate. Extra agent hours cost $3 each, metered per second.

The agent hour is the meter that matters. Chat, Research, Flows, Automate and custom agents all draw on it, and only time Quick spends actively working counts. Index storage beyond the pooled allowance is $5 per GB a month on every plan.

A worked band for a 50-person ops team on Enterprise: 50 × $40 plus $250 is $2,250 a month, about $27,000 a year, before overage. The same team on Professional pays $1,250 a month (about $15,000 a year) but loses dashboard authoring and any way to buy more hours. For comparison, Microsoft 365 Copilot lists at $30 per user/month on annual billing ($18,000 a year for 50 users), and Glean typically starts around $40-50 per user/month with a 100-seat minimum. Because the $250 fee is per account, Enterprise beats Max from five users up.

Best for

An ops or RevOps leader at a company that already runs on AWS and wants one tool for knowledge chat, recurring reports and BI, billed against the existing AWS commitment. It is the default move for any team that bought Amazon Q Business and needs a supported path forward.

Not for

A Microsoft 365 shop whose people live in Outlook and Teams: Microsoft 365 Copilot gets identity, distribution and in-document editing that Quick’s M365 extension does not match. It is also the wrong pick for a team that wants to build production agents in code; that is Amazon Bedrock territory.

Versus the alternatives

• Microsoft 365 Copilot — the largest installed base in enterprise assistants. Pick it when Entra ID, Teams and SharePoint are the centre of the company.
• ChatGPT Business or Enterprise — the other share leader, at $20 per seat on annual Business. Pick it when the priority is the strongest general assistant and the BI and automation can stay in other tools.
• Claude Team — the fastest-growing entrant, $20 per seat annual for standard seats. Pick it for heavy writing, analysis and coding work where model quality matters more than bundled BI.
• Gemini — pick it on Google Workspace, where it ships inside the plans you already pay for.
• Glean — pick it when permission-aware search across 100+ tools is the main job and the company has the seat count to clear its minimum.
• Tableau or Power BI — keep them when the BI estate is large and governed; Quick Sight is enough for ops dashboards, not for replacing a mature analytics team’s tooling.

Watch-outs

• Agent hours run out quietly on Professional. Professional includes 8 hours per user and cannot buy overage, so heavy Research or Automate use stalls. Guard: run the free 30-day trial (up to 25 users, Enterprise features), measure hours per user, and buy Enterprise for anyone above about six hours a month; cap overage with Limit Profiles.
• Self-serve plans sit outside your AWS account. Free, Plus and Max sign up with email or social login, so staff can put company data into a Quick account IT never provisioned. Guard: decide before rollout that company data goes only into Professional or Enterprise inside your AWS organisation, and say so in the AI acceptable-use policy.
• Q Business migration is not a lift-and-shift. Bring Your Own Index allows at most two Q Business indexes per region, Q Apps must be rebuilt as Flows, and guardrails and Actions do not carry over. Guard: inventory every Q App, guardrail and index before the move, and budget rebuild time per app rather than treating it as a licence switch.
• MCP tools time out at 60 seconds. Quick’s MCP client has a fixed 60-second timeout, no custom headers and a static tool list. Guard: wrap slow tools (CRM bulk queries, contract extraction) so they return a job ID quickly and post results back, and test each tool against the timeout before users depend on it.
• The BI promotion ends on 31 December 2026. Existing Quick Sight Author licences get Enterprise capabilities free until then. Guard: put the 2027 renewal in the budget now at Enterprise or Author Pro ($40) rates, not the $24 Author rate.

Related: the enterprise AI rollout stack shows where an assistant like Quick sits next to governance and model-platform layers.