OVHcloud · Tools

OVHcloud AI Endpoints

Visit website

At a glance

Price
Pricing on request
Vendor
OVHcloud

Specifications & properties

Key decision factors

Pricing model
Paid 1
Provider type
EU provider 1
Data residency
EU only 1
Certifications
  • ISO 27001
  • SOC 2
  • HDS (health data)
1
No training on customer data
Yes 1
Model catalogue (LLMs)
Llama, Qwen, DeepSeek and 40+ open-weight models 1

Integration

API available
Yes 1
OpenAI-compatible API
Yes 1
Enterprise SLA
Yes 1

Model

Open-weights models
Yes 1
Report data / suggest a correction

Serverless inference API from European sovereign-cloud pioneer OVHcloud, with 40+ open-weight models and a no-training, zero-retention guarantee.

Profile

OVHcloud AI Endpoints

OVHcloud AI Endpoints is a serverless inference API from OVHcloud, described as "a leading European provider of sovereign cloud services since 1999". It gives developers access to more than 40 open-weight models — including Llama, Qwen and DeepSeek — through popular APIs "like OpenAI", making it a low-friction European swap for US inference providers.

Data protection is a headline feature: "Your data will never be used to train or improve our AI models", and OVHcloud keeps "only the data required for billing purposes" (zero data retention). The platform holds ISO 27000-series, SOC and healthcare data (HDS) certifications and publishes a 99.5% availability SLA.

As a French operator with its own data centres, OVHcloud appeals to teams that want European jurisdiction plus breadth of open models. Within the EU AI cloud it competes with Scaleway and Nebius on token-based inference.

Frequently asked questions

Does OVHcloud train on my data?

No. OVHcloud states your data will never be used to train or improve its AI models, and it retains only what is required for billing.

Which models are available?

More than 40 open-weight models including Llama, Qwen and DeepSeek, reachable through an OpenAI-like API.