No. Vitruvian runs fully offline. In its enterprise, workstation, or server versions, no internet connection is needed.
However, an online version is available during the Public Beta phase — try it here → Vitruvian-1 Beta
For specific operational needs, even offline installations can be optionally extended with a controlled online search feature, allowing the system to query selected web sources while maintaining internal data isolation.

No. In standard installations, all data remain strictly on the notebook, workstation, or internal servers — nothing is transmitted outside the organization.
In the Public Beta version → Vitruvian-1 Beta, we temporarily store user prompts (questions only, not identities) for the sole purpose of improving core functionalities. These prompts are fully anonymized and not linked to any personal data.
If an organization requests custom model training, there are two options: the training can be done entirely on the client’s infrastructure, or ASC27 can handle it under secure, contractual agreement.

No. Vitruvian is not a general-purpose chatbot. It is designed for professional tasks, providing logical, structured, and task-specific answers rather than casual or conversational outputs.

Yes. Vitruvian supports local document ingestion, making it adaptable to specific operational or institutional needs.

Notebook (e.g., MacBook Air M2/M3, Dell XPS 14)
Workstation (e.g., Mac Studio, Dell Precision)
Server (e.g., HP, Lenovo, Dell models)
Minimum specs: 8 GB RAM with GPU

Vitruvian can be installed on your existing devices, or we can supply preconfigured machines ready to use.

Notebook: Best for individual use, small offices, mobile teams
Workstation: Suitable for technical, analytical, or intensive tasks
Server: Ideal for multi-user, centralized installations (e.g., managed by IT departments)

Yes. Vitruvian is free in its Public Beta online version, up to 1 million output tokens. Beyond that, a very low-cost usage model applies.
Try the Public Beta now → Vitruvian-1 Beta
For advanced scenarios, Premium and On-Premise licenses are available. → pricing

Yes, the Public Beta version is accessible online via browser → [link]
However, Vitruvian is built to work entirely offline in its full versions.
Optionally, it can include a module for online content search when required.

Vitruvian is a professional AI system developed by ASC27. It leverages existing foundational models through a customized AI pipeline, optimized for logical reasoning, structured outputs, and enterprise-grade usability.

ASC27 is an Italian Startup founded in 2017 in Rome, Italy. It focuses on applied AI for enterprise, defense, and cybersecurity sectors. It develops AI systems designed for full offline operation and on-premise deployment.

Yes. You can load internal reports, databases, manuals, and archives to allow the system to provide contextualized, knowledge-based answers. All processing remains local and secure.

Yes. ASC27 is a certified cybersecurity firm. Vitruvian integrates natively with SOC/CSIRT/CERT procedures and documentation to support operators in handling incidents, threats, vulnerabilities, reports, and more.

Yes. Vitruvian supports multilingual prompts and documents, including English, Italian, Spanish, French, and others. Language support can be extended on request for specific operational needs.

Yes. Vitruvian provides API endpoints and connectors that allow integration with internal software, ERPs, databases, and vertical tools. ASC27 also offers support for custom integration projects.

Yes. Fine-tuning or embedding optimization can be done on your internal corpus. Training can occur locally or under secure, contractual agreement with ASC27.

Yes. Vitruvian was designed specifically for secure public-sector deployment, including ministries, agencies, and local authorities, with on-premise and offline options available.