Pricing

Custom-Quoted On-Prem AI Infrastructure

Every deployment is sized around the inference capacity, hardware topology, connected applications and operational support your organisation actually needs.

01 · Quote foundation

Eight drivers. One clear scope.

We turn your infrastructure, inference and operating requirements into a custom proposal after discovery.

Quote scope
Infrastructure
Infrastructure footprint, node topology, accelerator and memory needs, networking, racking and distributor-backed procurement.
Inference
Inference estate, model mix, concurrency, context requirements, expected throughput and the cadence of capacity reviews.
Operations
Availability, update policy, recovery scope, support access and reporting define the recurring operating model.
02 · Product scope

A supported path from inference to API handoff.

The standard scope covers the inference core, recurring operations and a documented contract for the applications your teams already use.

Core Appliance
Local inference · operational control
The customer-premises foundation for private model serving, governed API access and day-to-day operations.
  • Local open-weight model inference
  • Model selection, configuration and validation
  • Documented Models API
  • Compact operational Console
  • Application credentials and model policies
  • Metadata-only activity and audit records
Managed Operations
Lifecycle · operational continuity
A recurring operating scope aligned to your availability, lifecycle, recovery and support requirements.
  • Observability layer and operational reporting
  • Agreed update and maintenance policy
  • Backup and recovery scope
  • Scheduled capacity reviews
  • Defined support access and escalation
  • Operational health and lifecycle management
API Handoff
Your applications · documented access
Chat Completions is the documented API protocol your favourite apps and harnesses use to access approved local models.
  • Application-specific credentials
  • Explicit model allowlists
  • Streaming Chat Completions
  • Tool-call message transport
  • Usage metadata in the Console
  • Validated API handoff and access documentation
03 · Deployment boundary

Inside your infrastructure.
Under your control.

LLM Machines provides private inference, recurring operations and a documented API contract. Your applications continue to access approved local models through that contract.

Infrastructure · 01

Customer-premises appliance.

The inference core, Console, identity, observability layer and lifecycle tooling are deployed inside the environment you control.

Data boundary · 02

Nothing leaves your infrastructure.

Inference inputs, outputs and workload content remain inside your environment, while operational records stay focused on metadata.

04 · Build vs. partner

The build-it-yourself maths.

What private on-prem inference can cost to assemble and operate internally, compared with a scope-based partnership. Illustrative Year 1 SME planning profile.

Option A · DIY

Build it in-house.

Hire the team. Stitch the OSS. Operate and maintain every layer.

  • 2× ML / infrastructure engineers (loaded)€220K
  • 1× DevOps / SRE€100K
  • 1× security and compliance lead€80K
  • Legal review: EU AI Act, GDPR, NIS2, EU Data Act€60K
  • Hardware procurement, racking and integration€100K
  • Security tooling, test capacity and model lifecycle work€36K / yr
  • Six months of integration, debugging and on-call workopportunity
  • Delivery and key-person riskunbounded
Year 1 typical €600K – €1M+
Option B · LLM Machines

Partner with us.

Define the operating scope without building and maintaining every platform layer internally.

  • Discovery and workload qualificationscoped
  • Distributor-backed hardware sizing and procurementscoped
  • Inference configuration and deployment validationscoped
  • Availability, monitoring and reportingagreed
  • Update, backup and recovery policyagreed
  • Capacity reviews, support access and escalationagreed
  • Documented API validation and operational handoffaccepted
  • Advanced integration deliveryseparate SOW
Commercial proposal Custom quote after discovery
Build in-house
€600K – €1M+
Deploy with LLM Machines
Scope-based proposal
The internal estimate uses illustrative loaded annual costs for a four-person platform, infrastructure and compliance team plus legal, hardware and tooling assumptions. Validate each input against your location and policies. LLM Machines pricing is provided only after discovery defines the deployment and recurring operating scope.
05 · Advanced integration consulting

A defined outcome beyond the standard API handoff.

When an integration needs more than documented model access, we scope a separate consulting engagement around one measurable operational outcome.

Outcome-led discovery
Problem · target · constraints
Define the workflow, systems, data boundary and target result before implementation begins.
  • Named business and technical owner
  • Bounded workflow and integration surface
  • Baseline and measurable target
  • Security and operating constraints
Separately scoped delivery
SOW · milestones · evidence
Customer application and workflow integrations are delivered under a separate statement of work, not assumed in standard onboarding or maintenance.
  • Explicit deliverables and exclusions
  • Milestone evidence and decision gates
  • Approved credentials and resource limits
  • Customer-selected applications and execution environment
Acceptance and handoff
Measure · document · own
Completion is tied to agreed acceptance criteria and a documented handoff into your team's ownership.
  • Measured result against the agreed target
  • Runbook and configuration handoff
  • Known limits and operating responsibilities
  • Customer ownership after acceptance
06 · Cloud vs. on-prem

Own the infrastructure and the operating boundary.

Move inference inside your environment while preserving the applications and workflows your teams already use.

Cloud AI APIs LLM Machines
Data controlPrompts leave your perimeterData stays inside your environment
Commercial modelUsage-based token billingCustom quote for owned infrastructure
PortabilityProvider-specific APIs and policiesDocumented API and source-available first-party Product source
DeploymentFast to start, external to your perimeterControlled deployment inside your infrastructure
AuditabilityLimited to provider exportsMetadata-only activity and audit records on-prem
07 · FAQ

Pricing questions.

How quotes, hardware, product scope and scaling work before a commercial conversation.

How is each deployment quoted?

We prepare a custom proposal after discovery. Scope drivers include the infrastructure footprint, inference estate, availability, update policy, recovery scope, capacity reviews, support access and reporting.

What is included in the standard product scope?

The standard scope combines the Core Appliance, operational Console, identity and policy controls, observability layer, lifecycle support and a documented API handoff.

How is Product source licensed?

Original first-party LLM Machines Product source is source-available under the unmodified PolyForm Internal Use License 1.0.0. Third-party software keeps its upstream licence.

Who owns the hardware?

The appliance is sized with your team, procured through distributor-backed channels and deployed as infrastructure you control. Warranty coverage follows the supplier and manufacturer documents and the signed Order Form.

How do existing applications connect?

Existing applications connect through the documented Chat Completions protocol and receive access only to approved local models.

How is integration consulting scoped?

Advanced integration consulting uses a separate statement of work with a measurable target, explicit acceptance criteria and a documented handoff into customer ownership.

What's next

Get a quote sized to your environment.

Discovery call, workload review, sized appliance specification and a quote matched to your environment.