fmcai.cloud
Sign inStart your project
Sovereign cloud · AI-native

Sovereign cloud & AI, in the region you choose.

A European-grade cloud and AI platform for teams worldwide. Your data and inference stay where you decide — France today, more regions on the same architecture. The boundary is enforced in code, not just promised.

Start your projectBuild with our team
Outside the CLOUD ActZero inbound portsHA failover, testedSovereign GPU inference
fmcai.cloud console — overviewfmcai.cloud console — overview, light theme
region · fr-gravelines-1 · your data stays here
Sovereignty · your jurisdiction

Sovereign to you — not to a datacenter.

You choose where your data and models live. The same architecture deploys per region, and workloads don’t leave the region you pick unless you explicitly opt in.

Client-defined residency

Pick the region your project runs in. Data, backups and inference are pinned there by design.

Outside the CLOUD Act

EU regions run on a French operator — no foreign jurisdiction reaching into your data.

Encrypted off-region backups

Disaster copies are encrypted at source; the key stays in your home region.

Zero inbound ports

Nothing is open to the internet. Your services are reached only through a private tunnel.

Regions
FR
France — Gravelines
fr-gravelines-1
Live
SG
Singapore
sg-singapore-1
On the roadmap
Your region?
same architecture, anywhere
Talk to us

France is in production today. Additional regions run the identical, reproducible stack — the sovereign boundary is technically enforced, per region.

Reliability

Built to stay up while it changes.

High availability across three physically separate machines. We’ve tested full chassis loss — the cluster kept quorum. Updates and patches roll through without taking your services down.

3×
physical quorum, across separate machines
0.240ms
private inter-node latency
tested
chassis-loss failover, not just claimed
0dt
maintenance with zero downtime
AI-native · preview

AI that stays where you do.

One OpenAI-compatible endpoint routes inference to sovereign GPUs — NVIDIA DGX, 128 GB unified memory. Your prompts and data never leave your region unless you pick a frontier model by name.

One endpoint, smart routing

Local sovereign GPU first; frontier models only on explicit, per-model opt-in.

Per-project keys & budgets

Every app gets a virtual key with its own quota, metered end to end.

inference.py
# Same code you already write for OpenAI —
# pointed at your sovereign region.
from openai import OpenAI

client = OpenAI(
    base_url="https://ai.fmcai.cloud/v1",
    api_key=FMCAI_KEY,   # per-project, scoped
)

resp = client.chat.completions.create(
    model="fmcai-sovereign-70b",   # stays in-region
    messages=[{"role": "user",
               "content": "Summarise last night's runs."}],
)
The platform

Everything a project needs.

Managed building blocks, provisioned for you. You ship code and manifests — compute, data, storage and security are already there.

Compute

Managed Kubernetes, one isolated namespace per project.

CNCF-certified

Managed Postgres

High-availability database with point-in-time recovery.

HA · PITR

Object storage

S3-compatible buckets with per-project keys.

S3 API

Cache

Low-latency key-value cache for your hot paths.

Redis-compatible

Secrets

Vaulted secrets injected at runtime — never in Git.

Per-app creds

Observability

Metrics, logs and traces included — no setup.

Automatic

Automatic TLS

Certificates issued and renewed at the edge.

Let's Encrypt

Sovereign AI inference

OpenAI-compatible endpoint on in-region GPUs.

Preview
FMC Management

You don’t have to build it alone.

FMC Management is more than infrastructure. We help you design and ship your project — then host and run it on the sovereign cloud. From idea to production, accompanied.

Build with us

Our team works alongside yours — architecture, development, delivery. Consulting or hands-on build, at your pace, until your project is real.

  • Solution & architecture design
  • Hands-on development & delivery
  • AI integration, done right

Host & run with us

We operate it on the sovereign cloud: provisioning, deployment, backups, monitoring and updates. You focus on the product, we keep it running.

  • Provisioning & GitOps deployment
  • Backups, monitoring, updates
  • Sovereign, in your chosen region
Pricing

Simple plans. One project or ten.

Start a single project or run a whole portfolio. No free tier — a real platform, priced fairly.

Starter
1 project
€49/mo
30-day free trial
  • 1 project · isolated namespace
  • Managed Postgres & storage
  • Observability included
  • Sovereign AI, base quota
Start free trial
Studio
5 projects
€199/mo
 
  • Up to 5 projects
  • Everything in Starter
  • 5× AI inference quota
  • Onboarding with our team
Choose Studio
Scale
Recommended
10 projects
€379/mo
 
  • Up to 10 projects
  • Everything in Studio
  • 12× AI inference quota
  • Priority accompaniment
Choose Scale
Custom
10+ projects
Let’s talk
 
  • Unlimited projects
  • Dedicated region options
  • Negotiated AI capacity
  • Dedicated accompaniment
Talk to us

No free tier · cancel anytime · prices excl. VAT · billed via Stripe

Start on a cloud that answers to you.

Spin up your first project in a sovereign region — or bring our team in to build it with you.

Start your projectBook a call