preorders open · shipping July 2026AI that fits
AI that fits
inside your office closet.
Turnkey AI workstations assembled in France, for running Llama 4, Qwen 3, DeepSeek, Mistral and 240 open-weight models — no cloud, no quota, no surprise bill.
preorder · build-to-order● assembled in France◆ 3-year Starter/Pro warranty
Pro example
3× RTX 5090
96 GB total VRAM
1800 W PSU
Models
242
+12 this week
GPUs
204
NVIDIA · AMD · Apple · Intel
Sync
24h
via Hugging Face
Cloud needed
0
never. by design.
Break-even
9mo
vs GPT-4o-class API
02 · cloud vs local
The math fits on one page.
simulation 1 agent · 8h/day · 30d
CloudGPT-4o-class API
€
Monthly cost
≈ 480 € — token billing
↗
Data leaving
to US servers — Cloud Act exposure
§
GDPR / AI Act
DPA, register, processor controls
~
First-token latency
120 — 800 ms depending on route
×
Quota / rate limit
RPM, TPM, queue during peak hours
∅
Models
≈ 8 in mainstream API catalogs
∞
Paid off
never · perpetual rent
LocalSTARTER rig · 1× 5090
€
Monthly cost
≈ 32 € — electricity
↘
Data leaving
none. stays inside your walls.
✓
GDPR / AI Act
You host and control the system.
⚡
First-token latency
8 — 40 ms on gigabit LAN
∞
Quota / rate limit
limited by your VRAM, not by a vendor
★
Models
242 · Llama, Qwen, Mistral, DeepSeek
9
Paid off
9 months on STARTER tier
verdict computed by /lib/calc.ts · build #2178net savings · 5 376 € / year / agent · break-even month 9
03 · rig catalog
Three chassis. One clear stance.
// 01 · Solo dev
Starter
4 990 €ex. VAT · delivered
- G1× RTX 509032 GB GDDR7 · 575 W
- R64 GB DDR5-6000Kingston Fury · CL30
- CRyzen 9 9950X16c · Zen 5
- S4 TB NVMeGen 4 + Gen 5
- ★198 / 242 models~110 t/s · Llama 8B Q4
popular
// 02 · Agency · Small team
Pro
11 990 €ex. VAT · delivered
- G3× RTX 509096 GB total VRAM
- R128 GB ECC DDR5RDIMM · 6000 MT/s
- CThreadripper 7960X24c · 5.3 GHz boost
- S8 TB NVMe Gen 5RAID-0 models
- ★234 / 242 models~38 t/s · Llama 70B Q4
// 03 · SME · Mid-market
Enterprise
25 990 €from · ex. VAT
- G2× RTX A6000NVLink · 96 GB · 112 GB/s
- R256 GB RDIMM ECC8 channels
- CTR Pro 7975WX32c Pro · 8-channel RAM
- S16 TB RAID-10NVMe Gen 5 · hot-swap
- ★Sovereign RAG includedSSO · RBAC · audit log
04 · what we stand by
An OS, not a kit.
Plug in. Serve.
▲
Sovereignty
GDPR and AI Act friendly. Your data stays in your room, measured and auditable.
◆
Bounded cost
Capex, not rent. Break-even in 9 months for typical agentic usage. No TPM ceiling.
●
Open stack
llama.cpp · vLLM · Ollama · Open WebUI · LlamaIndex · Qdrant. No lock-in.
■
Transparent math
VRAM estimates before quote. Real tokens/s will be added once the showroom rig is benchmarked.