How to roll Claude out to a non-technical team, the use cases that pay off first, and ROI benchmarks.
Claude for business
Claude Pro vs Team vs Enterprise: which plan, and is your data safe?
A plain-English guide to choosing between Claude's plans — and to the questions every business actually asks first: does Claude train on your data, where is that data stored (including for Australian businesses), and when should you use the API to meet compliance requirements.
The 30-second answer
- Pro is for individuals. Team (2–150 seats) is the right baseline for most small businesses — your data is never used for training by default.
- Enterprise adds SSO/SCIM, audit logs, a Compliance API and HIPAA-readiness for larger or regulated organisations.
- Anthropic does not train on Team, Enterprise or API data. On personal plans it may — unless you opt out.
- For data residency or strict compliance, use the API (Zero Data Retention, HIPAA/BAA) — and run it in an Australian cloud region if your data must stay onshore.
Plan comparison - Free, Pro, Max, Team and Enterprise — side by side.
The five Claude plans split cleanly into two groups: personal plans (Free, Pro, Max) and business plans (Team, Enterprise). The single biggest difference for a business is not price — it is how your data is treated.
| Feature | Free | Pro | Max | Teamfor business | Enterprisefor business |
|---|---|---|---|---|---|
| Price (USD) | $0 | US$17–20 / mo | From US$100 / mo | US$20–25 / seat / mo | US$20 / seat / mo + usage |
| Best for | Trying Claude | Individuals & power users | Heavy individual usage | Teams of 2–150 | Organisations & regulated industries |
| Users / seats | 1 | 1 | 1 | 2–150 seats | 20+ seats |
| Trains on your data? | Only if you opt in | Only if you opt in | Only if you opt in | No — off by default | No — off by default |
| Data retention | 30 days–5 yrs* | 30 days–5 yrs* | 30 days–5 yrs* | Per org policy | Custom retention controls |
| SSO & SCIM | — | — | — | SSO | SSO + SCIM |
| Audit logs & admin | — | — | — | Central admin & billing | Audit logs, Compliance API, RBAC, IP allowlisting |
| HIPAA-ready (BAA) | — | — | — | — | Yes |
*Consumer plans (Free, Pro, Max): data is kept ~30 days if you opt out of model training, or up to 5 years if you leave training on. Prices are Anthropic's US list prices, current as of October 2026. Viewed from Australia, Pro, Max and Team are listed in AUD with GST included (Pro A$34, Max A$170–340, Team A$34–43 per Standard seat). Anthropic changes plans often, so confirm the latest at claude.com/pricing before you buy.
Which plan for you - Match the plan to how you actually work.
Most businesses overthink this. Here is the short version by situation — and where the API earns its place over a subscription.
Start here — who's using Claude, and for what?
Just you
Pro
Full access to Claude and Claude Code for one person. Switch model training off if you handle client data.
Trains unless you opt outA team of 2–150
Team
Shared billing, SSO and organisation-wide search. The right baseline for most small businesses.
Never trained onRegulated, security-led or 20+ seats
Enterprise
Adds SCIM, role-based access, audit logs, a Compliance API and HIPAA-readiness with a BAA.
Never trained onBuilding a product or automating
API
Zero Data Retention, HIPAA/BAA and in-region (Australian) hosting — architecture you control.
Full controlFull feature-by-feature detail is in the comparison table above.
Your data & training - Does Claude train on your data?
This is the question that decides everything else, and the honest answer is: it depends on the plan.
Personal plans — Free, Pro, Max
Since a 2025 policy change, Anthropic may use your chats and coding sessions to train its models unless you opt out in Privacy Settings. Opt out and your data is retained ~30 days; leave it on and it can be kept up to 5 years. You can change the setting anytime, and deleted chats are not used for training.
Business plans — Team, Enterprise, API
Anthropic does not train its models on your content. The Commercial Terms put it plainly: “Anthropic may not train models on Customer Content from Services.” You keep rights to your inputs and own the outputs. This is the core reason any business handling client data should be on Team, Enterprise or the API — not a personal account.
Data residency - Where your data lives — and how Australian businesses keep it onshore.
For every first-party Claude product — the consumer apps, Team, Enterprise and the Claude API — data is stored at rest in the United States. On the API you can pin inference to US-only or allow global routing, but the storage region is US-only today. For many businesses that is perfectly fine. For some, it is not.
If your data must stay in Australia
Government, health, finance and many enterprise contracts require data to remain onshore. The path is to run Claude through a cloud provider that offers an Australian region rather than the first-party API:
- AWS Bedrock — Claude models in Asia Pacific (Sydney), ap-southeast-2, with Australia cross-region inference that keeps traffic within Australia (Sydney and Melbourne). Eligible AWS services carry IRAP assessments.
- Google Vertex AI — Claude models available in the Sydney region (australia-southeast1).
On these platforms the cloud provider — not Anthropic — is the data processor, so residency, IRAP and data-handling are governed by AWS or Google. Anthropic's own Zero Data Retention and HIPAA arrangements apply to the first-party API, not to Bedrock or Vertex, where you rely on the provider's equivalent controls.
API & compliance - When the API is the compliant choice.
A chat subscription is for people. When you have real compliance obligations — or you are building Claude into a product or workflow — the API is where the control lives. Six reasons it is the right tool for regulated work:
- It is never trained on. Anthropic's Commercial Terms state plainly that “Anthropic may not train models on Customer Content from Services.” You keep rights to your inputs and own the outputs.
- Zero Data Retention (ZDR). With a ZDR arrangement, Anthropic does not store your prompts or Claude's outputs after the response is returned — covering the core Messages API. Flagged-misuse content can be retained up to 2 years where law requires.
- HIPAA-ready with a BAA. Handling protected health information? Sign a Business Associate Agreement for HIPAA-ready API access. It is enforced at the org level, and non-eligible features are blocked automatically.
- Data residency you control. Run Claude through AWS Bedrock or Google Vertex AI in an Australian region and your data stays in-region, governed by that cloud provider's residency and IRAP controls.
- Audit logging & redaction. On the API you own the architecture: log every call, redact sensitive fields before they reach Claude, and gate actions behind human approval. Compliance becomes something you can evidence.
- Right-sized cost. Usage-based pricing with prompt caching and model routing (Haiku/Sonnet/Opus) often costs less than per-seat subscriptions for automated, high-volume workloads.
Security & certifications - What Anthropic is independently certified for.
Compliance is not just policy — it is audited. Anthropic maintains the attestations most procurement and risk teams ask for.
- SOC 2 Type I & Type II. Independent audit of security controls — Type II verifies they operated effectively over time. Reports available via Anthropic's Trust Center.
- ISO/IEC 27001:2022. The international standard for information security management systems (ISMS).
- ISO/IEC 42001:2023. The first international standard for AI management systems — specific to how AI is governed, not just IT security.
- HIPAA-ready (BAA available). For organisations processing PHI via the Claude API or Enterprise, with a signed Business Associate Agreement.
How Sandlabs helps - Pick the right plan — or build the compliant setup.
We are a Claude-first AI consultancy. Whether you just need the right plan switched on or a compliance-ready integration built on the API, we do the part that never happens on its own.
Not sure which plan — or whether you need the API?
Book a short strategy call and we will map your team size, data sensitivity and residency requirements to the right Claude setup — Team, Enterprise, or a compliant API build with Zero Data Retention and Australian-region hosting. If you simply want Claude switched on and working for your business, our done-for-you setup gets you there in one session.
FAQ - Claude for business: your questions, answered.
The questions we hear most from owners and operators weighing Claude up for their team.
- Does Claude train on your data?
- It depends on the plan. On consumer plans (Free, Pro, Max), since a 2025 policy change Anthropic may use your chats and coding sessions to improve its models unless you opt out — opt out and retention stays at ~30 days; leave it on and it can be kept up to 5 years. On commercial products (Team, Enterprise and the API), Anthropic does not train on your content by default. Deleted conversations are not used for training.
- What is the difference between Claude Team and Enterprise?
- Team is built for teams of 2–150: shared billing, SSO, organisation-wide search, up to US$500 a month in pooled API credits and no model training on your content by default. Enterprise is for larger or regulated organisations and adds SCIM user provisioning, role-based access control, audit logs, a Compliance API, custom data-retention controls, IP allowlisting and HIPAA-readiness with a BAA. Team is a flat per-seat price; Enterprise starts at 20 seats and is billed per seat plus usage at API rates.
- What is the difference between Claude Pro and Max?
- Both are personal, single-user plans with the same data policy - Anthropic only trains on your chats if you leave model training switched on. The difference is capacity: Pro (about US$17–20/month) gives one person full access to Claude and Claude Code; Max (US$100 or US$200/month) adds 5x or 20x more usage than Pro, higher output limits, priority access at busy times and US$100–200 a month in API credits. Max is for heavy individual use, not for sharing across a team - for that you want Team.
- Is Claude Pro or Team better for a small business?
- For a single operator, Pro is fine - but switch model training off if you handle client information. The moment a second person needs access, or you want a guarantee that your content is never used for training, Claude Team is the better fit: it now starts at 2 seats, training is off by default, it adds SSO and central admin, and a Standard seat is the same US$20 a month as Pro on annual billing.
- How much does Claude Enterprise cost?
- Claude Enterprise costs US$20 per seat per month, billed annually, with a 20-seat minimum, plus usage billed at standard API rates. You can buy it self-serve or through Anthropic's sales team, which adds invoicing on net terms, AWS Marketplace purchasing and committed-spend incentives. Admins can set organisation- and user-level spend limits. Because usage is billed separately, your real cost depends on how heavily your team uses Claude - talk to us and we will help you estimate it.
- How many users do you need for Claude Team or Enterprise?
- Claude Team now requires a minimum of 2 seats (down from 5) and scales to 150, and you can mix Standard and Premium seats. Claude Enterprise starts at 20 seats, either self-serve or through Anthropic's sales team.
- Do Claude Team and Enterprise include API access?
- Partly. A Team or Enterprise subscription gives your people access to Claude's apps - web, desktop, mobile and Claude Code. Team now also includes up to US$500 a month in pooled API credits; Enterprise does not. Building Claude into your own product, website or automation beyond those credits uses the Claude API, billed separately on a pay-as-you-go basis through the Console. Many businesses run both: a subscription for staff, and the API for the systems working behind the scenes.
- Should I use the Claude API or a subscription for my business?
- Use a subscription (Team or Enterprise) when people need to chat with Claude day to day. Use the API when you are building a product, automating a workflow, or have hard compliance requirements — it is the only path to Zero Data Retention, HIPAA-ready PHI handling, in-region data residency via a cloud provider, and architecture you control end to end. Many businesses run both: Team for staff, the API for the systems behind the scenes.
- Where is my Claude data stored, and can it stay in Australia?
- For all first-party Claude products — the consumer apps, Team, Enterprise and the Claude API — data is stored at rest in the United States. On the API you can pin inference to US-only or allow global routing, but the workspace storage region is US-only today. If you need data to stay in Australia, run Claude through a cloud provider with an Australian region: AWS Bedrock (Sydney, ap-southeast-2, with Australia cross-region inference) or Google Vertex AI (australia-southeast1). On those platforms the cloud provider is the data processor and residency is governed by their region and IRAP controls.
- Can Australian businesses use Claude under the Privacy Act?
- Generally yes, with the right setup — but meeting the Australian Privacy Principles (APPs) is your responsibility. Because first-party Claude stores data in the US, sending personal information to it is a cross-border disclosure under APP 8, so you should disclose it in your privacy policy, put a Data Processing Addendum in place, and take reasonable steps to ensure comparable protection. Where data must stay onshore, run Claude through an Australian cloud region (AWS Bedrock or Google Vertex in Sydney). This is general information, not legal advice — confirm your obligations with your privacy adviser.
- Is Claude HIPAA compliant?
- Anthropic offers HIPAA-ready access for the Claude API and for Claude Enterprise (sales-assisted), under a signed Business Associate Agreement (BAA). HIPAA-readiness is enforced at the organisation level and only covers eligible features. Consumer plans (Free, Pro, Max) and the Console/Workbench are not covered, so PHI workflows should be built on the API or run through Enterprise with a BAA in place.
- Is Claude GDPR compliant?
- Anthropic supports GDPR compliance for commercial customers: it offers a Data Processing Addendum (DPA), does not train on your content on commercial products, and holds ISO/IEC 27001 and SOC 2. Because first-party Claude stores data in the US, EU organisations with data-residency requirements typically run Claude through an EU cloud region (AWS Bedrock or Google Vertex in Europe) rather than the first-party API. As with any processor, GDPR compliance is shared — how you configure and use Claude matters too.
- What is Zero Data Retention (ZDR)?
- ZDR is an API arrangement where Anthropic does not store your prompts or Claude's outputs after the response is returned. It covers the core Messages API (and Claude Code via commercial keys or Enterprise) and is requested through Anthropic's sales team. Two caveats: it does not cover stateful features such as the Files or Batch APIs, and content flagged for policy violations can still be retained up to two years where the law requires. ZDR is the strongest retention guarantee available short of running Claude in your own cloud region.
- Is Claude safe for confidential or sensitive business data?
- Yes, with the right plan and setup. On Team, Enterprise and the API your content is not used for training, and Anthropic holds SOC 2 Type II, ISO/IEC 27001 and ISO/IEC 42001 certifications. For the most sensitive workloads, the API with Zero Data Retention (and an Australian-region cloud deployment where residency is required) gives you the strongest guarantees. Avoid putting confidential client data into a personal Pro account that still has model training switched on.
- What security certifications does Anthropic have?
- Anthropic holds SOC 2 Type I and Type II, ISO/IEC 27001:2022 (information security) and ISO/IEC 42001:2023 (AI management systems), and offers HIPAA-readiness with a BAA. Compliance reports can be requested through Anthropic's Trust Center.
Keep reading - Go deeper on Claude for business
Roll-out playbooks and cost breakdowns from the Sandlabs team.
Every Claude plan compared — Free, Pro, Team, Enterprise — with hidden costs and the right plan for your team size.
Plugins, Skills, MCP connectors — what each is, a decision tree, and worked examples for rolling Claude out to a team.
Per-model pricing, prompt caching, batch, and seven production tactics to cut Claude API spend.
Sources
- Claude pricing & plans
- Anthropic Commercial Terms of Service
- Anthropic Privacy Center (data use & retention)
- Claude API — data residency
- Claude API — data retention, ZDR & HIPAA
- Anthropic Trust Center
Last reviewed October 2026. Claude's plans, pricing and data policies change frequently — always confirm the current terms with Anthropic before making a compliance decision. Sandlabs is an independent consultancy. We are not affiliated with, endorsed by, or sponsored by Anthropic. “Claude” and “Anthropic” are trademarks of Anthropic PBC, used here only to describe the products we help businesses adopt.
Let's build something great together.
Melbourne, Australia — serving founders worldwide. [email protected]