Claude for business in Australia: Team vs Enterprise pricing in US and Australian dollars, GST, data residency, Privacy Act risks, plus a rollout plan.

Choosing Claude for business in Australia mostly comes down to one question: Team or Enterprise. For most organisations under about 50 staff, the answer is Team, on Standard seats with a few Premium seats for heavy users. Enterprise earns its higher real cost once you need automated user provisioning, audit logs or your own data retention rules. Neither plan keeps your data in Australia, and that is the part we’d spend the most time on before signing anything.

TL;DR: Claude Team costs US$20 per seat a month billed annually (US$25 monthly), about A$28 before any GST, with Premium seats at US$100. Enterprise is US$20 a seat plus metered usage at API rates, with a 20-seat minimum. Team suits most Australian SMBs; move to Enterprise when you need SCIM, audit logs or custom retention, and plan separately for data residency.

Australians already use Claude heavily: when Anthropic announced its Sydney office in March 2026, it said Australia ranks 4th globally in Claude.ai usage relative to population. The problem in most businesses is staff using it on personal accounts nobody manages, the shadow AI risk a company plan is meant to close.

Claude for business plans and prices in 2026

Prices checked on 2 November 2026 against claude.com/pricing, the Team plan page and the Enterprise plan page. Anthropic quotes US dollars and says “prices shown don’t include applicable tax”. Australian dollar figures use the Reserve Bank’s published rate for 30 October 2026 (A$1 = US$0.7040) and exclude GST.

PlanUS$ list priceApprox. A$ (ex GST)What you get
Pro (individual)$20/month, or $17/month billed annually~A$28 / ~A$24One person, Claude Code included
Max / Ultra (individual)$100 / $200 per month~A$142 / ~A$2845x and 10x Pro usage
Team, Standard seat$25 monthly, $20/month billed annually~A$36 / ~A$282 to 150 seats, SSO, central billing, Claude Code
Team, Premium seat$125 monthly, $100/month billed annually~A$178 / ~A$142Five times the usage of a Standard seat
Enterprise$20/seat/month billed annually, plus usage at API rates~A$28 per seat, plus usage20-seat minimum, SCIM, audit logs, Compliance API, custom retention

Pro, Max and Ultra suit a sole trader, not a team: each person controls their own account, and model training is opt-out rather than off by default.

A ten-person firm with eight Standard seats and two Premium seats on annual billing pays US$360 a month, roughly A$511, or about A$6,140 a year before GST. The smallest Enterprise contract, 20 seats at US$20, is US$4,800 a year (about A$6,820) before a single token of usage is billed.

What about GST?

Anthropic’s help centre says your billing address determines how tax is calculated and that a Tax or VAT ID field appears when your address is eligible, with changes applying only to future invoices. We couldn’t find anything from Anthropic on how it treats an Australian ABN. Under the ATO’s rules, GST on imported digital services applies to sales to Australian consumers, and a GST-registered business buying for business use generally isn’t one. Enter your ABN under Settings, then Billing, before the first invoice if the field appears, check whether 10% was added, and ask your accountant how to record it.

What Team actually includes

Team is the plan most Australian businesses should start on. Every seat includes chat, Projects, connectors, memory, Claude Cowork and Claude Code; Anthropic’s Claude Code help article says “Claude Code is included with every Team plan seat.” SSO and domain verification are included, and there is no model training on your content by default.

Seat type matters more than people expect, because Cowork and Claude Code eat into the same seat allowance as chat. Standard seats suit staff who draft, summarise and research. Premium seats are for the two or three people running long agent tasks or coding all day. You can mix seat types, add seats mid-term, and let an owner switch on extra usage when someone hits a limit. Team’s pooled API credits (up to US$500 a month) can’t be spent in Claude Code or the Claude apps, so don’t count them as spare capacity.

What Team lacks is the governance layer: Anthropic’s comparison table marks SCIM, audit logs, role-based access, the Compliance API and custom data retention as “No” for Team. If you’re running Claude Cowork, our Claude Cowork guide covers the safety settings that product needs.

When Claude Enterprise is worth it

The Enterprise seat looks the same price as Team Standard, but the model is different. You pay US$20 a seat and then your team’s actual consumption at API rates, which Anthropic says “scales with model and task”. Light users can come out cheaper than on Team; heavy agent work can cost far more. Get an estimate from a month of real Team usage before signing the annual contract.

That uncertainty buys you four things Team doesn’t have.

  • SCIM provisioning and domain capture, so leavers lose access when they leave your identity provider, not when someone remembers.
  • Audit logs, OpenTelemetry monitoring and a Compliance API for programmatic access to activity logs, chats, files and projects.
  • Custom data retention. Owners can set retention down to a 30-day minimum; without a setting, data is kept indefinitely. It covers chats, projects and artifacts, but not Managed Agents or other features that run Claude Code in the cloud.
  • Role-based access, IP allowlisting and a HIPAA-ready configuration with a BAA.

One old argument for Enterprise no longer holds. The pricing page lists the same context window for Pro, Team and Enterprise: up to 1M tokens, varying by model.

Our rule of thumb: move once you’re past 20 to 30 users and someone needs evidence of who did what, or a client contract asks how long you keep their data. Below that, Team plus a written AI usage policy covers most of the risk.

Data handling, residency and the Privacy Act

Anthropic says it will not use inputs or outputs from commercial products to train models by default. The exception is feedback: a conversation a user rates with thumbs up or down may be used, and Anthropic’s retention article says feedback data is kept for five years. Owners can switch feedback off with the “Rate chats” setting, and we’d do that on day one. Deleted conversations leave Anthropic’s back-end within 30 days.

Residency is where expectations run ahead of reality. The data residency controls on Anthropic’s developer platform offer only “global” or “us” inference, with “us” the only storage location. In March Anthropic said it was “exploring adding local capacity through our third-party partners in Australia”, a statement of intent rather than a product. Today no setting on Team or Enterprise keeps your chats in Australia.

If you need Australian processing now, the route is Amazon Bedrock, which offers Claude through an Australian cross-Region inference profile. Called from Sydney or Melbourne, requests stay in those two Regions; called from Auckland, they can also run in New Zealand. Check which Claude models the profile covers before you plan around one. Anthropic’s Claude Desktop on third-party platforms runs the desktop app on top of a provider like that, for organisations that “cannot send conversation data to the United States”, and keeps conversation history on the user’s device. It is managed through device management or an admin console against your own cloud account: an IT project, not a plan upgrade.

Sending client personal information to an overseas provider can engage APP 8 of the Privacy Act. You must take reasonable steps to ensure the overseas recipient doesn’t breach the APPs, and section 16C can leave you accountable if it does. Read the OAIC’s guidance on commercial AI products, then run Anthropic through our AI vendor data privacy checklist before personal information goes near it.

Claude vs ChatGPT Business and Microsoft 365 Copilot

Claude is the stronger choice when your business runs on Google Workspace or a mix of tools, when the work is long documents, analysis and drafting, and when some staff also write code, since Claude Code comes in every Team seat.

It is the weaker choice in two situations. If you live in Outlook, Teams and SharePoint, Microsoft 365 Copilot works inside the compliance boundary you already manage, and Microsoft’s April 2026 update expects local inferencing for Copilot interactions in Australia by the end of 2026, a date that has already moved once from an end-of-2025 target. And if Australian storage is a contractual requirement, ChatGPT Enterprise has offered data at rest in Australia for new workspaces since November 2025, Computerworld reported, though inference still runs in the US. Our Microsoft 365 Copilot guide and Copilot vs Gemini vs ChatGPT comparison cover those options.

A rollout plan that sticks

Having led digital programs for organisations including NSW Government, UNSW Sydney and Telstra, I’ve learned that the licence is the easy part of any software rollout. With AI tools, the hard part is deciding what staff may paste in, and that decision is the one most often skipped.

  1. Week 1: rules first. Agree which data classes are allowed, write them into your AI policy, and find out who already uses personal Claude or ChatGPT accounts.
  2. Week 2: set up Team. Verify your domain, turn on SSO, enter your ABN, switch off “Rate chats”, approve connectors one at a time, and cap extra usage.
  3. Weeks 3 to 6: pilot. Give Standard seats to five to ten people across two teams, Premium seats to your heaviest one or two users, and a named workflow each. Measure time saved.
  4. Week 7: decide. Expand Team, adjust the Premium mix, or open the Enterprise conversation with real usage data.

If you’d rather not run it alone, our AI tools and automation setup service, part of our AI and automation practice, can do the setup with you and leave a documented handover.

The verdict on Claude for business

For most Australian SMBs, Claude for business means a Team plan at about A$28 a Standard seat before GST, with Premium seats only where usage proves the need. The policy on what data goes in matters more than the seat mix. Enterprise is worth it when governance evidence becomes a requirement, not before. If Australian data residency is non-negotiable, plan a Bedrock-based deployment or look hard at the alternatives.

If you’d like help choosing the right plan and rolling it out properly, book a free 30-minute call. If you’d rather prove the value first, our AI Quick-Win Audit is an A$2,950 flat-fee, ten-business-day engagement that ends with a live demonstration of one automated workflow in your business.

Frequently Asked Questions

How much does Claude Team cost in Australia?

Anthropic prices Claude Team in US dollars: US$20 per Standard seat a month billed annually or US$25 monthly, and US$100 or US$125 for a Premium seat. At the RBA rate for 30 October 2026, a Standard annual seat is about A$28 before any GST.

What is the difference between Claude Team and Claude Enterprise?

Both include SSO, Claude Code and no training on your content by default. Enterprise adds SCIM, audit logs, role-based access, a Compliance API and custom data retention, and bills usage at API rates on top of a US$20 seat, with a 20-seat minimum.

Is Claude Code included in the Team plan?

Yes, in every Team seat. It uses the same seat allowance as chat and Cowork, so people coding all day often need a Premium seat.

Does Claude store data in Australia?

Not on Anthropic’s own plans, which offer no Australian storage option today. Organisations needing Australian processing can use Claude on Amazon Bedrock through its Australian inference profile, optionally with Claude Desktop on third-party platforms.

Does Anthropic train on my business data?

Not by default on Team, Enterprise or the API. Conversations a user rates with thumbs up or down may be used, so owners should switch off the “Rate chats” setting.