LocallyAI

Installed pricing

The whole system.
A clear starting price.

Hardware, the Harness, local models, setup and handover are priced together. We start with the work your team needs done, then quote the memory, storage and platform that can comfortably handle it.

Software and hardware in one quote · Australian dollars · prices exclude GST · GST registered · valid 14 days

Five useful starting points

Pick by workload, not headcount alone.

Two people working through long files can need more memory than a larger team doing short drafts. These tiers make the first conversation easier; the quote names the real configuration.

Micro

One person, dictation and drafting

from $2,900 AUD

A 16 GB Mac mini M6 running the Fast lane and the embedding model only. The right answer when the work is transcription, document search and drafting for one person, and nothing larger is needed.

  • Apple M6, 16 GB unified memory
  • 512 GB fitted storage
  • Fast lane and embedding model — no second model resident
  • Harness, setup and handover included

Compact

One or two regular users

from $4,900 AUD

A quiet Mac mini for core chat, document search, transcription and everyday drafting.

  • Apple silicon, normally 32 GB memory
  • 1 TB fitted storage
  • Fast and Smart model lanes sized to the machine
  • Harness, setup and handover included

Office

A small shared team

from $7,900 AUD

More memory and room for shared use, with a Mac or AMD Strix Halo configuration.

  • Mac mini M5 Pro, or Strix Halo for Windows/Linux
  • 48–64 GB memory, quoted to the workload
  • 1 TB fitted storage
  • Shared matters, files, tools and staff accounts

Practice

Heavier models or more work at once

from $10,900 AUD

Mac Studio, 128 GB Strix Halo or NVIDIA DGX Spark, chosen around speed, operating system and software needs.

  • 64–128 GB memory
  • 2 TB fitted storage
  • Mac, Windows or Linux options
  • CUDA option with DGX Spark

Firm

High memory, concurrency or paired systems

from $17,500 AUD

A Mac Studio M5 Ultra, multiple appliances or paired DGX Spark systems for the jobs that genuinely need them.

  • 96 GB and above, quoted to the model set
  • 4 TB fitted storage
  • Department or workload separation where useful
  • Network and backup plan included in the scope

Hardware choice

Why the same tier can contain different machines.

Mac mini or Mac Studio
our usual recommendation for speed and ease of use
AMD Strix Halo
Windows or Linux, 64–128 GB, strong value
NVIDIA DGX Spark
Linux, CUDA and 128 GB coherent memory
Multiple appliances
more simultaneous work or a deliberate department boundary

Sized to the work, not to the ceiling.

These are planning tiers, not automatic recommendations. If a 16 GB Mac running a small model and transcription passes the agreed workload, we will quote that instead. The right machine is the one that matches the work — not the largest one we could sell you.

The names on a tier are choices, not a bundle of every machine. Your quote contains one exact configuration unless it clearly says otherwise. See the current hardware and trade-offs →

The three LocallyAI hardware paths side by side: Apple Mac Studio, HP Z2 Mini Strix Halo and NVIDIA DGX Spark
Mac — recommended · Strix Halo — Windows or Linux · DGX Spark — Linux and CUDA

Included

What the installed price covers.

Computer and fitted storage
included
LocallyAI Harness
included
Local model set
selected and configured
Document search and transcription
set up and checked
User access and basic permissions
configured
Initial templates and folders
fitted to the agreed scope
Handover
staff walkthrough and setup record

Separate when needed

Costs that are not hidden in the base.

Optional yearly visit
$600 per year
Custom work
$200 per hour, quoted first
Complex integrations
fixed scope before work starts
Off-site fine-tuning
quoted after a data review
Staff accounts
no per-user Harness fee; machine capacity still applies
GST
Registered; GST is shown separately on the written quote and invoice

Ongoing costs

No meter running on every conversation.

The local models do not create a per-message cloud bill. Electricity, storage, backups and any separately enabled online service still have ordinary costs. If an integration has its own vendor fee, it stays that vendor's fee. You purchase the hardware rather than renting a seat; your accountant should determine its tax and accounting treatment.

Updates

A short connection, not standing access.

Applying an update needs a temporary connection while files are fetched and checked. Improvements are included in the support plan rather than billed again, and release notes make clear what changed.

Before a quote

We need three honest examples.

Show us a typical short job, a heavy job and the awkward one that usually breaks the process. Use real examples with sensitive details removed; they become the acceptance set the delivered system must pass.

We are ready to start a closed pilot of the platform with businesses. On-site installs and demonstrations are coming soon; register your interest now.

Register interest for a pilot See the install process

Own the working system

Why businesses choose LocallyAI over ChatGPT or Claude.

A LocallyAI quote buys configured hardware, Harness software, tested workflows and handover. Headcount does not create another Harness licence charge; compare ownership cost over the same period as recurring per-seat fees.

Connections are quoted, not assumed

Each supported integration, permission boundary and acceptance test is written into the quote. It is not treated as delivered until it passes on the final machine.

Quoted automations run on the appliance

A morning brief, weekly digest or aged-debt list is included only when named in the quote and proven against agreed examples before handover.

Processing location is the difference

Hosted services require your business to rely on a provider’s infrastructure, terms, configuration and continuing service. LocallyAI keeps approved base workflows on your appliance; connected exceptions are documented and acceptance-tested.

A boundary your team can inspect

The standard workflow is local. Web research, integrations, remote support, updates and off-site backups are separate exceptions recorded in the quote and tested before handover.

Core work can continue offline

Local chat, approved document search and transcription can keep running when the connection drops. Web research and connected services cannot.

Built from scratch, for this

We did not rebrand a coding interface. The Harness is written for office workflows and fitted to the templates, letterhead and approved systems named in your quote.

The Harness is modular

Supported skills and MCP connections can be added without replacing the interface. Each requested module still needs a written scope and delivery test.

You can actually talk to us

We build the software, so a requested change can be discussed with the people who write it. Standard product improvements and business-specific work are separated in the written scope, including any cost.

Pricing questions

The bits worth checking.

Shared improvements are not billed as bespoke work.

If a requested improvement belongs in the standard Harness, the quote can treat it as a platform change rather than customer-specific development. Business-specific integrations, rules and deliverables are scoped and priced separately.

What is included in the starting price?

The selected computer, storage, the LocallyAI Harness, a fitted set of local models, document search, transcription setup, user access, basic templates, installation and handover. The written quote lists the exact hardware and any work outside that base.

Are there per-user licence fees?

We do not charge a Harness fee for each person. The practical limit is the amount of work the machine can handle at once, so a growing team may eventually need a larger or second appliance.

Is the yearly visit compulsory?

No. The optional $600 yearly visit is a scheduled health check and tune-up. Support and quoted update work are separate from that visit, and your machine does not stop because you skip it.

Why are quotes valid for 14 days?

Memory and workstation prices move. A 14-day window lets us name the exact machine and hold a realistic number. Once accepted, the quote governs the agreed work.

Do Harness updates need the internet?

Applying an update needs a temporary connection. The normal local setup can return offline after the update is checked. We do not need standing remote access.

How is GST shown?

Published prices exclude GST. LocallyAI is registered for GST, so GST is shown separately on the written quote and invoice.

Can we supply our own machine?

Yes. Bring us the machine and we will assess what it can genuinely run — memory, storage, operating system, and how much headroom is left once a model is loaded.

If it suits the work, the hardware line comes off the quote entirely and you pay for the software and configuration only: the Harness, the local models, your own templates and the setup. If it is not up to the job, we will say so plainly and tell you what it would take, rather than install something that will disappoint you.

Bring us one job your team repeats every week. We will show you the system doing it →