Develop your web apps with AI, in France
An AI-powered full-stack web application generator powered by bolt.diy, installed and maintained by DINAO. Describe, run, edit, and deploy — with the LLM of your choice, on your French servers.
What is bolt.diy?
bolt.diy is the official open-source version of Bolt.new: a tool that allows you to describe, run, edit, and deploy full-stack web applications directly in the browser, all driven by AI. You state your need in natural language, and bolt.diy generates the code, runs it, and lets you adjust it.
Its key strength is being model-agnostic: you choose the LLM used for each prompt from a wide range of providers — OpenAI, Anthropic, Ollama, OpenRouter, Gemini, LM Studio, Mistral, DeepSeek, Groq, Cohere and many others — and the project easily extends to any model supported by Vercel's AI SDK.
Maintained by StackBlitz Labs and a large open-source community, bolt.diy is self-hosted and becomes a complete AI prototyping workshop that you keep under your own control.
Host bolt.diy at DINAO
Resource tiers compatible with bolt.diy prerequisites (minimum 1 vCPU / 1 Go / 2 Go). Hosted in France, fully managed.
- 1 dedicated vCPU
- 2 Go RAM
- 20 GB NVMe
- Daily backups
- Managed & monitored by DINAO
- 2 dedicated vCPU
- 4 Go RAM
- 40 GB NVMe
- Daily backups
- Managed & monitored by DINAO
- 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
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 might be wondering…
Which AI models can I use?
bolt.diy is agnostic: OpenAI, Anthropic, Mistral, Gemini, DeepSeek, Groq, Cohere… or a local model via Ollama / LM Studio. You choose the LLM for each generation.
Does the generated code remain private?
Yes. In local mode, your code and prompts stay on your French instance. In external API mode, prompts pass through the chosen provider.
Where is the data hosted?
On DINAO infrastructure in France, in one of the available data centers. In local mode, your data does not leave the territory.
Are technical skills required?
Web knowledge helps to exploit the generated code, but description-based prototyping makes the tool accessible. DINAO handles installation, security, and updates.
Can I export the generated applications?
Yes. The produced code belongs to you and remains fully exportable, without proprietary lock-in.