Skip to main content

Generative AI, without sending your data outside of France.

Dedicated GPUs in France, billed by the hour of operation
Models running on your GPUs, or the Mistral API hosted in France
RAG for your CAD, PDM, and documentation data, without exposing it
Request a Callback
View the AI offer

4.9/5 on Google1,500+ customersNumber one 3DEXPERIENCE reseller worldwide in 2025

The Seesaw

Using AI shouldn't mean giving away your assets

Most generative AI services send your requests to servers outside Europe. For industrial or regulated data, this is a real risk, not just a hypothetical scenario.
01

The CLOUD Act and U.S. APIs

Your prompts and documents pass through extraterritorial jurisdictions. Hosting in Europe is not enough if the operator is governed by U.S. law.
02

Intellectual Property on Display

CAD drawings, specifications, contracts, code: sending this intellectual property to a third-party model means losing control over your most sensitive data.
03

Dependency and Costs

Pay-as-you-go billing, opaque pricing, and lock-in to a proprietary API. An open model built on your infrastructure makes costs predictable and allows for easy exit.
What We Cover

From models to operations, AI stays with you

Six building blocks that work together: from the model deployed as a sovereign API to the assistant specialized in your data, with an engineer managing the whole system.
01 · Deployment

Models that run on your GPUs

Mistral, Llama, and other open-weight models deployed for private inference on your sovereign GPU machines. Your weights, your infrastructure, no external calls.
02 · Sovereign API

Mistral models hosted in France

Get started quickly: access to Mistral models hosted by Outscale Dassault Systèmes on SecNumCloud-certified infrastructure, with data never leaving France.
03 · RAG

A helper for your business data

Connect the model to your data sources: CAD/PDM, technical documentation, standards, and contracts. Answers based on your data, not the web.
04 · Adaptation

Specialization and Agents

Train an open model on your corpus, equip agents on your internal systems, all within your sovereign environment.
05 · Managed Services

A dedicated engineer

GPU sizing, deployment, security, cost and version tracking. One point of contact, not a support ticket.
06 · Safety

Segregation and Compliance

Private network, encryption, isolation by project, and optional SecNumCloud region for the most sensitive data.

Un cas d'usage en tête ?

Décrivez-le, on vous dit ce qu'il faut pour le faire tourner sur une infrastructure en France.

Request a Callback
Technical Evidence

A sovereign inference stack, from the GPU to the model

The GPU is billed by the hour it's powered on: we turn it on for jobs and turn it off the rest of the time. That's the whole point of the hourly pricing model for AI.
Sizing example
INFERENCE · SOVEREIGN STACKlive
COMPONENTS
GPUNvidia A10 · L40 · A100 · H100
RuntimevLLM · private inference
OrchestrationOutscale Managed Kubernetes (OKS)
Regioneu-west-2 · cloudgouv (optional)
GPU
1 to 8 cards
Context
up to 128k
Billing
on time
Private Network & VPNEncryptionInsulation by ProjectSaving TemplatesSupervisionReversibility
Use Cases

AI to Support Your Technical Roles

01

Research Assistant

Query standards, procedures, PLM documentation, and specifications using natural language.
02

Design Office Assistant

Support the design process using your CAD/PDM data, without exposing it to a third party.
03

Document Analysis

Extract, summarize, and compare contracts, requests for proposals, and technical reports.
04

Enhanced Technical Support

Respond more quickly to customers and teams using your knowledge base.
05

Secure In-House Assistant

A tool for handling sensitive data in the defense, healthcare, and public sectors.
06

Business Automation

Agents configured on your internal systems to streamline repetitive tasks.

Préparez votre demande en quelques questions.

Quelques questions sur votre besoin, puis un spécialiste revient vers vous.

Launch the simulator
Where to Start

Four Entry Points to Your Sovereign AI

Standard configurations to quickly define the scope. The exact scope (model, GPU, use case) is finalized during the quotation process. Everything starts with a scoping exercise, never with an online purchase.
Driver

Discovery

To validate an initial use case.
✓Lightweight model, served via a sovereign API or on a GPU VM on an hourly basis
✓RAG on a data source
✓A few users
✓Guided Setup
Production

Team

For a production assistant.
✓Balanced model, dedicated GPU
✓Multi-source RAG
✓IT Outsourcing and Cost Tracking
✓Standard Region or SecNumCloud
Sovereign

Advanced

For intensive and gentle use.
✓Self-hosted template
✓GPU Cluster, High Availability
✓Enhanced insulation
✓Optional SecNumCloud region
Review

Custom-made

For a regulated or specific need.
✓Specialization and Agents
✓Integration with Your IT System
✓SecNumCloud
✓Dedicated Framing
Le chiffrage se fait au cadrage, selon le modèle, le GPU et la volumétrie. Le GPU se facture à l'heure allumée.
Why Ohmycad

A partner, a network, support

A pioneer of Dassault Systèmes offerings, Ohmycad makes SOLIDWORKS, CATIA, and Outscale accessible, without the complexity.

1 500+
clients supported
#1
Number one 3DEXPERIENCE reseller worldwide in 2025
600+
companies supported with Ohmycare
4.9/5
based on 311 verified Google reviews
Accessibility
A partner who knows you
A network of local partners, Ohmycad Solution Partners and Expert Partners, fully dedicated to Dassault Systèmes solutions.
Agility
A model that adapts to your business
Our plans adapt to your needs: you start with what you need and scale up as your business grows. No over-provisioning.
Customer Experience
A dedicated engineer for your team
A dedicated, certified engineer will provide you with ongoing support to help you get the most out of your licenses, ensuring you never run into any roadblocks.
Expertise
The right expert, at the right time
An international network of certified experts: the best expertise on any topic, whenever you need it.
They trust usAcusurgicalAispridBiossunCalomatechEHMMcKay BrothersPonticelli BrothersAerix Systems

Vos données restent en France.

Parlons de votre projet IA et de ce qu'il demande côté infrastructure.

Request a Callback
Customer Reviews

They explain it better than we do

4.9 /5

311 Google reviews

Read all customer reviews View the Google listing
Mohammad Aghababa a year ago

I had a great technical support session with Alan and Léo from Ohmycad. They explained clearly how to save files on 3DEXPERIENCE and manage revisions with the maturity cycle. The session was very helpful, professional, and easy to follow. Looking forward to the next one on automating the drawing title block. Highly recommend their support team!

Pala Innovations 10 months ago

Alan, Celine and the rest of the team are extremely helpful. Instant support and great customer experience! highly recommended.

Benoit Lefort a year ago

As a startup working with Solidworks, we’re truly grateful for the brilliant support we've been receiving from Ryan and Hady. Their guidance, their clarity, and their ability to adapt to our stakes make a real difference. This kind of team by your side is a huge asset when moving fast and building deeptech. Thanks guys !

Jules Humpheson 2 years ago

Excellent support, XDI respond very quickly in a positive, friendly manner and have been able to resolve the few problems I've had, very quickly and efficiently. I would not hesutate to recommend them.

Mohammad Aghababa a year ago

I had a great technical support session with Alan and Léo from Ohmycad. They explained clearly how to save files on 3DEXPERIENCE and manage revisions with the maturity cycle. The session was very helpful, professional, and easy to follow. Looking forward to the next one on automating the drawing title block. Highly recommend their support team!

Pala Innovations 10 months ago

Alan, Celine and the rest of the team are extremely helpful. Instant support and great customer experience! highly recommended.

Benoit Lefort a year ago

As a startup working with Solidworks, we’re truly grateful for the brilliant support we've been receiving from Ryan and Hady. Their guidance, their clarity, and their ability to adapt to our stakes make a real difference. This kind of team by your side is a huge asset when moving fast and building deeptech. Thanks guys !

Jules Humpheson 2 years ago

Excellent support, XDI respond very quickly in a positive, friendly manner and have been able to resolve the few problems I've had, very quickly and efficiently. I would not hesutate to recommend them.

Mohammad Aghababa a year ago

I had a great technical support session with Alan and Léo from Ohmycad. They explained clearly how to save files on 3DEXPERIENCE and manage revisions with the maturity cycle. The session was very helpful, professional, and easy to follow. Looking forward to the next one on automating the drawing title block. Highly recommend their support team!

Pala Innovations 10 months ago

Alan, Celine and the rest of the team are extremely helpful. Instant support and great customer experience! highly recommended.

Benoit Lefort a year ago

As a startup working with Solidworks, we’re truly grateful for the brilliant support we've been receiving from Ryan and Hady. Their guidance, their clarity, and their ability to adapt to our stakes make a real difference. This kind of team by your side is a huge asset when moving fast and building deeptech. Thanks guys !

Jules Humpheson 2 years ago

Excellent support, XDI respond very quickly in a positive, friendly manner and have been able to resolve the few problems I've had, very quickly and efficiently. I would not hesutate to recommend them.

Mohammad Aghababa a year ago

I had a great technical support session with Alan and Léo from Ohmycad. They explained clearly how to save files on 3DEXPERIENCE and manage revisions with the maturity cycle. The session was very helpful, professional, and easy to follow. Looking forward to the next one on automating the drawing title block. Highly recommend their support team!

Pala Innovations 10 months ago

Alan, Celine and the rest of the team are extremely helpful. Instant support and great customer experience! highly recommended.

Benoit Lefort a year ago

As a startup working with Solidworks, we’re truly grateful for the brilliant support we've been receiving from Ryan and Hady. Their guidance, their clarity, and their ability to adapt to our stakes make a real difference. This kind of team by your side is a huge asset when moving fast and building deeptech. Thanks guys !

Jules Humpheson 2 years ago

Excellent support, XDI respond very quickly in a positive, friendly manner and have been able to resolve the few problems I've had, very quickly and efficiently. I would not hesutate to recommend them.

Questions About Sovereign AI

Sovereign AI, Explained

Can I use Mistral without sending my data outside of France?
Yes. Mistral models are either deployed on your sovereign GPUs in France or accessed via an API through the Outscale Dassault Systèmes infrastructure, which is SecNumCloud-certified. In both cases, your data and prompts remain subject to French law.
Should we self-host the model or use an API?+
Both options are possible. Self-hosting (open-source software running on your GPUs) maximizes control and cost predictability; the sovereign API simplifies setup. The choice depends on your data volume and constraints.
Can sovereign AI be classified as SecNumCloud?+
Yes, as an option. Outscale Dassault Systèmes operates a cloudgouv region certified as SecNumCloud 3.2 by ANSSI. It is enabled for projects that host particularly sensitive data.
Is my data used to train the models?+
No. Your data is used to respond to your requests and, if you choose, to fine-tune a model for your own use. It is not shared with any third-party training models.
How many GPUs do I need for my project?+
It depends on the model, the number of users, and the target latency. We start by sizing the system based on a specific example, then fine-tune it. The GPU is billed by the hour it’s running: we start small, shut it down between jobs, and scale up when the workload demands it.

Let's put a number on your sovereign AI project

✓An Overview: Use Cases, Data, Volume, Compliance
✓Transparent Pricing: GPU by the Hour, Model, Services, and Managed Services
✓A dedicated engineer, from deployment to operations
Request a callback →Overview of the Offer