An AI server for business: from the user to the compute, inside your network.
MagnolIA Server serves the models from hardware installed in your network. Each piece has a job; only monitoring goes outside.
MagnolIA Server is the AI server for business that runs open-weight models on hardware installed in your network. It identifies each user, applies their permissions and quotas, distributes work across machines, logs usage and exposes an API compatible with OpenAI's and Anthropic's.
- Compatible API
- Endpoints compatible with the OpenAI API and the Anthropic API: Claude Code, Cursor or Cline work by changing only the server address and the key.
- Audit log
- A chained, verifiable audit log, retained for two years.
- Backups
- AES-256 encrypted backups that allow the service to be restored on another machine.
- Capacity
- Capacity is added or removed without interrupting the service and without anything changing for users.
From the user to compute, without leaving your network
The service at a glance. The orchestration layer is our key piece: it lets us grow and maintain everything with no impact on your users.
Your teams' access
People sign in through a secure chat and developers via API, each with their own identity, under your internal domain.
Encrypted entry
A single encrypted access point, inside your network. None of the system is exposed to the outside.
Orchestration
Our IPThe heart of our service: it identifies each user, distributes the work and measures usage. This is where our know-how lives.
Local compute
The models run on the equipment installed in your network and respond in real time. Grow by adding capacity.
Traceability & control
Per-user usage logging and role-based access, audit-ready. Only monitoring travels outward.
What the platform provides
One service for people, developers and agents. Everything runs in your network and follows the same identity, permission and logging rules.
Standard-compatible APIs
OpenAI- and Anthropic-compatible endpoints. The tools your team already uses —Claude Code, Cursor, Cline or your own integrations— work by changing only the address.
Text, code and reasoning
State-of-the-art open-weight models, selected and maintained by us, with task profiles: fast answers, code, writing or in-depth analysis.
Speech and transcription
Multilingual audio and meeting transcription with speech models running on your infrastructure.
Documents and images
PDF and image understanding, plus embeddings for semantic search and RAG over your internal documentation.
Private web search
A dedicated search engine that gives models access to current information without sending queries to third-party services.
Anonymisation
Personal and sensitive data replaced in documents before processing, with the original data restored afterwards.
Intelligent load balancing
Requests go to the machine that can answer soonest, keeping each conversation on the same node to reuse its cache. Capacity is added or removed without interrupting service.
Identity, roles and quotas
Users, groups and companies with per-model and per-machine permissions, usage quotas, scoped API keys and two-factor authentication for administrators.
Audit and continuity
A hash-chained, verifiable audit trail retained for two years, and encrypted backups that allow the service to be restored on another machine.
The hardware, unbranded
From a compact node to a high-performance rack. The exact hardware behind each Núcleo is decided and managed by us based on your usage and volume; you only see the capacity. We renew it without you noticing.

Compact node
- capacidad
- ~18 users
Performance rack
- capacidad
- ~40 users
High-performance rack
- capacidad
- ~100 users
Maximum-capacity rack
- capacidad
- ~250 users
From your team's workstation to local compute
Select a use case to adapt the workstations and the AI's work, and scale the Núcleo to see how hardware is added with nothing changing for your users.
Use case
- 1
Team workstations
Developers · Operations / Support · Product / Business
- 2
Encrypted access
chat + API · TLS
- 3
Our IP
Orchestrating app
- 4
Núcleo M
Performance rack · GPU ×2 · ~40 users
- 5
Monitoring
the only outbound link
Use case
Software company
TECH
AI work
- ›Agentic coding
- ›RAG over the codebase
- ›Support automation
Capacity
~40 users
Núcleo M · GPU ×2
Scale your power
We move up capacity live: we add hardware to the Núcleo with nothing changing for your users. Everything is included: equipment, management, support and continuous model improvement.
The client only consumes. We operate.
- ›
Users only interact with the chat or the endpoint; never with the system, the models or the infrastructure.
- ›
For the user it's a black box: no technical access, no admin panel. Operations are exclusively ours.
- ›
Each user only sees their own: nobody accesses others' conversations, data or sessions.
- ›
Isolated network: compute and internal services are not exposed; only the authenticated access point is reachable.
Turnkey, in phases
- P1
Infrastructure
Installation of the equipment and base, inside your network.
- P2
Models
Tuning the models for your case.
- P3
Integration
Connection with your systems and first tests.
- P4
Pilot
Secure access and a pilot with a group of 10–20 people.
- P5
Production
Whole org, corporate access, monitoring and backups.
Frequently asked questions about MagnolIA Server
What is MagnolIA Server?
MagnolIA Server is the software that serves AI models from the hardware installed in your network. It identifies each user, applies their permissions and quotas, distributes work across machines and logs usage. It exposes an API compatible with OpenAI's and Anthropic's for chat, code, speech, documents and search.
Which AI models does MagnolIA use?
MagnolIA uses state-of-the-art open-weight models, which it selects, tests and updates, with profiles per task: quick answers, code, writing or in-depth analysis. When a better model appears, we bring it in ourselves, without your team having to do anything or switch tools.
Is it compatible with the tools my team already uses?
Yes. MagnolIA Server exposes endpoints compatible with the OpenAI API and the Anthropic API. The tools your team already uses, such as Claude Code, Cursor, Cline or your own integrations, work by changing only the server address and the key.
What can the platform do besides chat?
Beyond text, code and reasoning, MagnolIA Server transcribes audio and meetings in several languages, reads PDFs and images, generates embeddings for semantic search and RAG over your documentation, anonymises documents and offers private web search. Everything runs on your infrastructure under the same identity, permission and logging rules.
What happens if my team grows?
If your team grows, capacity is added without downtime: MagnolIA adds hardware to the Núcleo without interrupting the service and without anything changing for users. The server distributes requests across the available machines and keeps each conversation on the same one to make use of its cache.
Does it integrate with company single sign-on and with a SIEM?
Yes. MagnolIA's corporate mode includes sign-in with the company account (SSO) and organisation-wide policy. On the server there is audit output for the SIEM, model routing, document collections and encrypted backups. Everything comes switched off until the organisation itself enables it.



