Claude AI packages

Claude packages for South African businesses.

Two ways to put Anthropic's Claude to work: a fixed-price Desktop setup for one person, or a scoped rollout for your team. Set up by a studio that runs its own business on it every day.

Packages

Pick the shape that fits.

Claude Desktop, personal
R5,000 once-off
One person. Claude installed, configured and working on your real tasks by the end of the week.
  • Plan chosen with you (Pro or Max) and account set up in your name
  • Claude Desktop installed on your Mac or PC, plus web and mobile
  • Your email, calendar and files connected through the official connectors
  • Three projects built around your actual work, with instructions and files
  • A prompt library for your role: drafts, summaries, research, admin
  • Memory and preferences set so it writes the way you do
  • Two hours of one-on-one training on a call, or at your desk if you are in Johannesburg
  • 30 days of email support after handover
Book a Desktop setup

Excludes the Anthropic subscription, billed to you directly in US dollars.

Claude for Business
Quoted after a mapping call
Teams of 3 to 25 that want Claude doing named tasks under rules, with someone accountable for the result.
  • Mapping session: the two or three tasks that eat the most time
  • Claude Team or Enterprise rollout, seats, billing and admin
  • Roles, usage policy and a written set of standing rules
  • Shared projects and a company knowledge base
  • Connectors and integrations: email, calendar, drive, CRM, accounting, analytics
  • Service accounts per system, scoped and revocable
  • Guardrails: drafts not sends, ask before anything destructive
  • Two weeks supervised, every correction turned into a rule
  • Team training and a written playbook
  • Optional: Claude Code on your servers for monitoring, reporting and fixes
  • Monthly review and support retainer
Book a mapping call

Scope and price depend on how many systems Claude needs to reach.

The personal setup is a fixed R5,000. Business rollouts are quoted in rand after a short mapping call, because the cost is in the setup and supervision, not the software. Anthropic bills the subscription directly, so you are never locked in through us.

What you actually get

AI that does the work, not a chatbot in the corner.

Most "AI for business" offers end at a login and a webinar. Ours ends with Claude doing named tasks in your business, with rules, with someone accountable for the result.

In our own studio Claude reads server logs and proposes fixes, pulls monthly reports from Google Analytics, Search Console and Ad Manager through their APIs, loads advertising bookings, drafts client correspondence in our tone and keeps notes on every project. It never sends, deletes or restarts anything without a person saying so. That is the deployment we build for you.

Read how it looks day to day in What a Claude deployment actually looks like for a small business.

How a deployment runs

Five steps for a business rollout.

01

Map the work

A two-hour session to find the repetitive, judgement-light tasks that eat your week: reporting, first-draft correspondence, data cleanup, monitoring, documentation. We pick two or three to start.

02

Wire the access

Service accounts and API keys, read-only where possible, scoped per system. No shared passwords, nothing with more reach than the task needs.

03

Write the rules

Tone, what it may send, what it must ask about, what it must never touch. These become its standing instructions and we test them against real cases.

04

Run it supervised

Two weeks of Claude doing the work with a person checking every output. Every correction becomes a rule. You see exactly what it gets right and wrong before it runs alone.

05

Measure and extend

Hours saved, errors caught, turnaround time. If the numbers are not there we stop. If they are, we add the next task.

The rules that make it safe

Guardrails are the product.

Separate identities

One service account per system, per client. A key labelled for one purpose never touches another. Everything is revocable in one place.

Drafts, not sends

Email, invoices and anything outward-facing is drafted for a human to send. Autonomy is widened one task at a time, once earned.

Destructive actions wait

Deleting, restarting, dropping: Claude stops, shows the evidence and asks. It is written into its standing instructions, not left to chance.

Your data stays yours

Sensitive data lives on infrastructure you control. Anthropic does not train on business or API data. Claude sees the task, not the filing cabinet.

Memory with discipline

Claude keeps notes on every project: what was done, what broke, what not to repeat. Those notes are reviewed and corrected like any colleague's.

Measured, or stopped

Hours saved, errors caught, turnaround times. We report them monthly. If a task is not paying for itself we take it off Claude's plate.

Questions we get asked

Straight answers.

What does a Claude deployment cost in South Africa?

The personal Desktop setup is R5,000 once-off, excluding the Anthropic subscription, which you pay directly in US dollars. Business rollouts are quoted after a mapping call and depend on how many systems Claude needs to reach; most are two- to three-week projects with an optional monthly retainer. No surprises after the quote.

Is our data safe?

Anthropic does not train on business and API data, and we keep sensitive data on infrastructure you control. Claude sees what it needs for the task through scoped accounts, not your whole filing cabinet. We document every access path and you can revoke any of them in one place.

Do we need a developer on staff?

No. Both packages are built for owner-run businesses and small teams without an IT department. If you want Claude operating servers as part of a Business rollout, someone technical needs to own them, and that can be us.

Which Claude plan should we be on?

For individuals, Pro covers most knowledge work and Max suits heavy daily use. For teams, Claude Team gives shared projects and admin controls; Enterprise adds single sign-on and audit logs. We choose with you in the first session and you pay Anthropic directly, so you are never locked in through us.

Can Claude send email or make changes on its own?

Only if you decide it should, and we advise against it at the start. Our default rules are drafts not sends, and ask before anything destructive. Autonomy is widened task by task once the supervised period shows it is earned.

What happens after handover?

Business rollouts include a monthly review where we look at what Claude did, what it got wrong and what to add next. Desktop clients get 30 days of email support and can book more time afterwards.

Next step

Start with a mapping call.

Forty-five minutes, no charge. You tell us what eats your week, we tell you which package fits and what it will cost. If Claude is not the right tool for the job we will say so.

Book a mapping call

Claude and the Claude logo are products of Anthropic. Chronon provides independent deployment, integration and support services. See also our Claude AI deployment overview.