Hatch Your Stack

Choose the capacity that fits your business.

Every Hatched Stack uses the same business-first operating-system approach. Choose based on the capacity you need for local models, knowledge, development, and simultaneous services.

Hatched Stacks configurations

Launch, Pro, Max, or Ultra.

Every configuration begins with one carefully scoped workflow. Final hardware, integrations, controls, responsibilities, and acceptance criteria are confirmed after consultation.

Starting point

Hatched Launch Stack

$9,997

Mac mini · 16 GB / 512 GB

For entrepreneurs and growing teams starting with one cloud-first production workflow.

  • One production workflow
  • Cloud-first · light local models · limited simultaneous services
  • Up to 3 workflow connectors
  • Business context and approved model routing
  • Human approvals, activity records, and recovery documentation
Choose Launch Stack
More capacity

Hatched Pro Stack

$11,997

Mac mini · 24 GB / 1 TB

For established operators and medium-sized businesses that need more capacity and expansion room.

  • One production workflow
  • Moderate local capacity · larger knowledge base · greater concurrency
  • Up to 5 workflow connectors
  • Business context and approved model routing
  • Human approvals, activity records, and recovery documentation
Choose Pro Stack
Maximum Mac mini capacity

Hatched Max Stack

$13,997

Mac mini · 48 GB / 1 TB

For medium-sized businesses preparing to expand AI workflows across business functions.

  • One production workflow
  • Stronger local inference · heavier development · higher concurrency
  • Up to 8 workflow connectors
  • Business context and approved model routing
  • Human approvals, activity records, and recovery documentation
Choose Max Stack
Want to run a completely local Hatched Stacks AI operating system?Explore the Mac Studio option for approved local-processing workflows.
Completely local AI processing

Hatched Ultra Stack

From $29,997

Workload-selected Mac Studio

For approved business workflows requiring completely local AI model processing.

  • One approved local-processing workflow
  • Local model evaluation · private knowledge environment · workload engineering
  • Defined by accepted scope
  • Workflow-specific data boundaries and processing-path documentation
  • Final architecture selected after workload qualification
Choose Hatched Ultra Stack →

The next step is a short qualification application, not payment. Final configuration, scope, acceptance criteria, and purchasing terms are confirmed after consultation.

Included in every accepted build

A governed foundation, not a box of disconnected tools.

Business knowledge

Organized company, offer, customer, workflow, policy, and project context that the system can use with explicit boundaries.

Model routing

Task-appropriate routing across approved cloud or local models, with customer-owned accounts and disclosed processing paths.

Agent operating layer

Specialized roles, task handoffs, checkpoints, approvals, and exception handling instead of disconnected prompts.

Production workflow

A scoped workflow can be observed, baselined, built, shadow-tested, and considered for production under agreed acceptance criteria.

Governance and reliability

Permissions, activity records, spending controls, health checks, logs, backups, recovery procedures, and rollback paths.

Control and continuity

The accepted scope defines access, account ownership, documentation, training, backup, recovery, and any stabilization support.

After the initial engagement

Evolve support is optional.

Unless the accepted scope says otherwise, ongoing maintenance and support are separate. Evolve plans may include monitoring, maintenance, advisory support, and a bounded workflow sprint. Fit and monthly pricing are disclosed after a stack consultation.

Without an Evolve plan, the customer owns ongoing development, upkeep, model and API usage, subscriptions, backups, and operational maintenance.

Not included

Boundaries protect the build.

  • Unlimited workflow development or an unlimited whole-business transformation
  • Model and API usage, advertising spend, and third-party subscriptions
  • Regulated professional work or uncontrolled high-risk actions
  • Guaranteed cost savings, revenue, profit, or performance
  • Indefinite maintenance, support, or workflow expansion

Frequently asked questions

Important boundaries before you apply.

What exactly is the AI operating system?

A dedicated environment that organizes business context, routes approved models, connects workflow-specific tools, coordinates agent roles, enforces approvals, records activity, and includes operating and recovery documentation.

How are timing and deliverables defined?

We propose a scoped first-workflow plan after qualification. The accepted order defines timing, dependencies, deliverables, access, tests, support, and acceptance. Website descriptions are targets and examples, not automatic commitments.

How do I choose the machine configuration?

Compare Launch, Pro, and Max above based on the capacity your workload needs. Expand Hatched Ultra for approved workflows requiring completely local AI model processing.

Does local-only mean every connected service is private and local?

No. Local-only applies only to the approved AI-processing workflow and its documented data path. External websites, CRMs, email, payments, and other authorized services operate under their own data practices.

Do I have to buy monthly support?

No. You can self-manage, use your own technical team, commission fixed-scope work later, or choose a Hatched support plan after consultation.

Why does the engagement cost more than the Mac?

The Mac is one component. Depending on the accepted scope, the engagement can include discovery, business-context organization, system design, configuration, implementation, integrations, controls, testing, documentation, training, and stabilization support.

Founding capacity is limited

Three paid founding builds.

Pricing will be reviewed after 60 operating days. The current formal review date is October 1, 2026.

Build My Stack
Hatch Your Stack · Hatched Stacks