Professional machines engineered for specific domains.
Where personal Stera ends, Forge begins.
A personal Stera machine is a general-purpose sovereign intelligence — it handles the breadth of your daily work. Forge machines are different. They are purpose-built for specific professional domains, with hardware configurations, pre-installed models, and Talent Pod selections tuned for the demands of that field.
A research lab needs sustained inference on large models and the ability to fine-tune on proprietary datasets. A design studio needs vision models that understand creative tools and can generate at production resolution. An engineering floor needs real-time inference with deterministic latency. These are not the same workload. They should not run on the same machine.
Every Forge machine runs SteraOS. Every Forge machine is sovereign — your data stays local, your models are yours, no subscription, no cloud dependency. What changes is the hardware, the pre-installed intelligence, and the professional extensions that make it exceptional in its domain.
Forge Research Forge Design
┌─────────────────────┐ ┌─────────────────────┐
│ GPU: 48-128 GB VRAM │ │ GPU: 24-48 GB VRAM │
│ Focus: Large models │ │ Focus: Vision models │
│ Fine-tuning │ │ Generation │
│ Long context │ │ Multi-modal │
│ │ │ │
│ Talent Pods: │ │ Talent Pods: │
│ Research Analyst │ │ Visual Designer │
│ Literature Review │ │ Brand Strategist │
│ Data Scientist │ │ Content Creator │
│ Hypothesis Engine │ │ Asset Generator │
└─────────────────────┘ └─────────────────────┘
Forge Engineering Forge Medical
┌─────────────────────┐ ┌─────────────────────┐
│ GPU: 24-48 GB VRAM │ │ GPU: 24-48 GB VRAM │
│ Focus: Code models │ │ Focus: Clinical NLP │
│ DevOps │ │ HIPAA-ready │
│ Real-time │ │ Encrypted │
│ │ │ │
│ Talent Pods: │ │ Talent Pods: │
│ Senior Developer │ │ Medical Scribe │
│ QA Lead │ │ Clinical Researcher│
│ DevOps Engineer │ │ Billing Coder │
│ System Architect │ │ Patient Liaison │
└─────────────────────┘ └─────────────────────┘
Each Forge workstation is a complete professional AI system for an individual or small team. It sits on your desk or in your lab. It comes pre-configured with the models and Talent Pods optimized for your domain — ready to work on day one.
Forge Server
┌────────┐ ┌────────┐ ┌────────┐ ┌────────┐
│ A6000 │ │ A6000 │ │ A6000 │ │ A6000 │
│ 48 GB │ │ 48 GB │ │ 48 GB │ │ 48 GB │
└────────┘ └────────┘ └────────┘ └────────┘
Unified GPU Pool: 192 GB VRAM
SteraOS AI Center
┌───────────┐ ┌───────────┐ ┌───────────┐
│ Model │ │ Inference │ │ Pod │
│ Depot │ │ Engine │ │ Roster │
└───────────┘ └───────────┘ └───────────┘
┌───────────┐ ┌───────────┐ ┌───────────┐
│ Joules │ │ Training │ │ RBAC │
│ Economy │ │ Engine │ │ + SSO │
└───────────┘ └───────────┘ └───────────┘
│
┌──────────────┼──────────────┐
│ │ │
┌────┴─────┐ ┌────┴─────┐ ┌────┴─────┐
│ Terminal │ │ Terminal │ │ Mobile │
└──────────┘ └──────────┘ └──────────┘
When a domain needs more than a single workstation — when a research group, a legal department, or a clinical team needs centralized AI infrastructure — the Forge Server provides multi-GPU compute for an entire team.
Software Talent Pods run on the server. Team members connect through Stera Terminals at their desks or Stera Mobile pads in the field. The server is pre-configured for your industry — compliance frameworks, domain models, and professional Talent Pods are installed before delivery.
A Talent Pod is a specialized AI worker — a model fine-tuned for a specific professional role, packaged with the occupation knowledge and tool integrations that role requires. Talent Pods are what give Forge machines their domain expertise.
Software & Engineering
─────────────────────────────────────────────────
Senior Developer CodeForge 32B 24 GB
QA Lead CodeForge 32B 24 GB
DevOps Engineer DevOps 7B 8 GB
System Architect Architect 14B 16 GB
Data Analyst DataPilot 7B 12 GB
Research & Science
─────────────────────────────────────────────────
Research Analyst Research 14B 16 GB
Literature Reviewer Research 14B 16 GB
Data Scientist DataPilot 14B 16 GB
Hypothesis Engine Research 32B 24 GB
Healthcare
─────────────────────────────────────────────────
Medical Scribe Medical 7B 12 GB
Clinical Researcher Medical 14B 16 GB
Billing Coder Medical 7B 12 GB
Patient Liaison Support 7B 12 GB
Legal & Compliance
─────────────────────────────────────────────────
Contract Analyst Legal 14B 16 GB
Compliance Officer Compliance 7B 12 GB
Legal Researcher Legal 14B 16 GB
Paralegal Legal 7B 12 GB
Creative & Design
─────────────────────────────────────────────────
Visual Designer Vision 14B 16 GB
Copywriter Creative 7B 12 GB
Content Strategist Creative 14B 16 GB
Brand Strategist Creative 14B 16 GB
Finance
─────────────────────────────────────────────────
Financial Analyst Finance 7B 16 GB
Risk Assessor Finance 14B 16 GB
Compliance Analyst Compliance 7B 12 GB
Talent Pods can be installed on any Forge machine — workstation or server. On a workstation, they run directly on the local GPU. On a server, they run as software pods in the shared GPU pool. You can add, remove, or swap Talent Pods at any time without affecting the rest of the system.
Custom Talent Pods can be co-developed with your organization — trained on your proprietary data, calibrated to your domain terminology, integrated with your specific tools and workflows.
Forge Workstations
─────────────────────────────────────────────────
Forge Research Lab
GPU 48 GB (A6000) — 128 GB (GB10)
CPU 12-core Ryzen 9 / Threadripper
RAM 128-256 GB DDR5
Storage 2-4 TB NVMe
Training On-device fine-tuning capable
Price from €8,999
Forge Design Studio
GPU 24 GB (RTX 5090) — 96 GB (2× A6000)
CPU 12-core Ryzen 9 / Threadripper
RAM 64-256 GB DDR5
Storage 1-4 TB NVMe
Vision Multi-modal perception
Price from €4,299
Forge Engineering
GPU 16 GB (RTX 5070 Ti) — 48 GB (A6000)
CPU 12-core Ryzen 9 / Threadripper
RAM 64-128 GB DDR5
Storage 1-4 TB NVMe
Latency Sub-100ms inference
Price from €3,299
Forge Medical
GPU 24 GB (RTX 5090) — 96 GB (2× A6000)
CPU 12-core Ryzen 9 / Threadripper
RAM 64-256 GB DDR5
Storage 1-4 TB NVMe (encrypted)
Compliance HIPAA pre-configured
Price from €4,999
Forge Server
─────────────────────────────────────────────────
Standard 4× A6000 = 192 GB VRAM
High-perf. 4× H100 = 320 GB VRAM
Ultra 8× H200 = 1.1 TB VRAM
CPU Dual Xeon / EPYC
RAM 256 GB — 2 TB DDR5 ECC
Storage 8-32 TB NVMe RAID
Network 10/25 GbE
Users 20-500+
Terminals Stera Terminal €199-349 each
Price from €35,000 (contact us for
H100 / H200 configurations)
Common to All Forge Machines
─────────────────────────────────────────────────
Operating Sys. SteraOS (pre-installed)
Talent Pods Pre-configured for domain
Monthly fees €0
Per-token costs €0
Data residency 100% on-premise
Subscription None — you own the hardware
The machines listed above are starting configurations. Every Forge deployment can be customized — different GPU configurations, different storage capacities, different Talent Pod selections, different compliance pre-configurations. For organizations that need maximum compute density, Forge Servers are available with NVIDIA H100 and H200 GPUs — up to 1.1 TB of unified VRAM in a single system. If your domain is not listed, we will build the right Forge machine for it.
We also offer Forge bundles — a workstation for each team member plus a Forge Server for the department, pre-configured to work together from day one. The workstations handle individual work. The server handles heavy inference, training, and shared intelligence.
Tell us your domain, your team size, and the work you need AI to do. We will design the right Forge configuration and deliver it ready to operate.