The vector database for your AI, hosted in France
A vector database powered by Milvus, installed and maintained by DINAO. Your embeddings and similarity indexes remain on our French servers — the sovereign foundation for your AI and RAG applications.
What is Milvus?
Milvus is an open-source vector database designed to store, index, and query embeddings at massive scale. It is one of the most widely used similarity search engines in the world for building AI applications: retrieval-augmented generation (RAG), chatbots, semantic search, recommendation, duplicate detection, or image search.
Milvus offers a wide range of indexes (HNSW, IVF, Flat, DiskANN) and distance metrics (L2, cosine, dot product), with optional GPU acceleration. It supports hybrid search combining vector similarity with scalar filters, and exposes gRPC/REST APIs as well as Python, Go, Java, and Node SDKs.
Its cloud-native architecture separates compute and storage, allowing for a standalone deployment to get started or a distributed (cluster) setup to reach a billion vectors. It integrates natively with modern AI frameworks (LangChain, LlamaIndex, Haystack) to become the search memory for your language models.
Host Milvus at DINAO
Resource tiers compatible with Milvus prerequisites (minimum 2 vCPU / 8 Go / 20 Go disque). Hosted in France, fully managed.
- 4 dedicated vCPU
- 8 Go RAM
- 80 GB NVMe
- Daily backups
- Managed & monitored by DINAO
- 8 dedicated vCPU
- 16 Go RAM
- 160 GB NVMe
- Daily backups
- Managed & monitored by DINAO
2 hidden tier(s) (insufficient resources for this app) : Découverte, Standard
This application uses AI
The container hosts the application, not the AI engine (which requires dedicated GPUs) : inference is handled externally, with your own provider key. Prioritize a sovereign engine — Mistral AI, NumSpot, Scaleway, or OVHcloud AI Endpoints (France, GDPR) ; an international provider (OpenAI, Anthropic…) only if a specific capability requires it. Inference subscriptions are not included in hosting.
Technical details
You may also like…
Other applications managed by DINAO in the same category, hosted in France and GDPR compliant.
You might be wondering…
What is a vector database?
It is a specialized database for storing and searching 'embeddings' — numerical representations of text, images, or audio. It finds the most semantically similar items, which is essential for RAG and similarity search.
Does Milvus replace my LLM?
No, it complements it. Milvus is the search memory: it provides your LLM with relevant passages (RAG). You keep your chosen model, whether local or remote.
Where are my vectors hosted?
On DINAO infrastructure in France, in one of the available data centers. Your embeddings do not leave the country.
What volume can Milvus handle?
From a few million vectors in standalone mode to several billion in distributed deployment, with response times in the millisecond range.
Can I change plans or export my data?
Yes. You can upgrade or downgrade at any time, and your collections remain exportable — no proprietary lock-in.
