OVHcloud announces the official launch of AI Endpoints, a new serverless cloud solution designed to facilitate the integration of artificial intelligence models into business applications and services. Created to democratize the use of AI, the platform offers a library of over 40 open source models, covering a wide range of professional use cases.
AI Endpoints allows developers to access models for natural language processing (LLM), code generation, speech recognition, and image generation without managing the underlying infrastructure or having extensive machine learning expertise. The solution is based on OVHcloud's sovereign cloud environment, ensuring secure data hosting in Europe and protection against non-European regulations.
The platform offers a sandbox environment for developers to experiment with model capabilities before full integration, as well as APIs that facilitate integration into existing development environments. Targeted use cases include customer service automation, coding assistance, text extraction from unstructured documents, and voice-to-text and text-to-voice conversion. OVHcloud emphasizes the transparency of its approach, with models provided in open weight, which can be deployed on third-party infrastructures or on-site according to organizational needs.
After a preview, the solution has been progressively enriched with new models, taking user feedback into account (support for stable models, refined API key management, etc.). Currently, AI Endpoints offers more than 40 state-of-the-art open-source AI models, including:
LLM: Llama 3.3 70B, Mixtral 8x7B, ...
SLM: Mistral Nemo, Llama 3.1 8B, ...
Code: Qwen 2.5 Coder 32B, Codestral Mamba
Reasoning: DeepSeek-R1 (Llama distilled)
Multimodal: Qwen 2.5 VL 72B, ...
Image Generation: SDXL
Voice and Speech: ASR (speech-to-text), TTS (text-to-speech)
The platform allows on-demand inference, with usage-based pricing based on the number of tokens consumed per minute and per model. Hosted at the Gravelines data center near Dunkirk, which uses a water cooling system to minimize environmental impact, it is already available in Europe, Canada, and the Asia-Pacific region.

Cet article publirédactionnel est publié dans le cadre d'une collaboration commerciale

Translated from L'IA clé en main : OVHcloud lance AI Endpoints, sa plateforme serverless d’IA open source

To better understand

What does on-demand inference mean as used by OVHcloud's AI Endpoints platform?

On-demand inference allows for the flexible and optimized use of resources by processing data only when a request is made, which can reduce costs and improve energy efficiency.

How does OVHcloud's sovereign cloud comply with European data protection regulations?

OVHcloud's sovereign cloud ensures all data remains in Europe, adheres to GDPR standards, and provides protection against unauthorized access by extra-European regulations, ensuring data confidentiality and sovereignty.