We bring AI infrastructure inside your company. One local platform for private chat, RAG, agents, document intelligence and internal APIs — without a single document leaving your network.
We are not selling a PC with a graphics card. I3K Local AI is a complete platform — validated hardware, EuLLM as the inference runtime, RAG Enterprise for document intelligence — that runs artificial intelligence inside the company perimeter. Models, prompts, documents and answers stay on the customer's systems. No tokens metered, no data crossing the boundary.
Most companies today pay for five or six separate AI services, each with its own contract, its own per-token price and its own data transfer. I3K Local AI covers these workloads locally.
Inference runs on the hardware on site. No per-token cost, no billing surprise when usage grows. The cost is the hardware plus the software subscription.
The appliance is purchased by the customer, not leased. It stays a company asset even if you change software vendor later on.
We work with AMD, NVIDIA and other accelerators we have validated on our stack. If you have an explicit CUDA requirement, there is a dedicated configuration.
Start with a single appliance and grow toward multi-GPU or multi-node when concurrency demands it, without redesigning the architecture.
Prices exclude VAT, are indicative and subject to confirmation. Every configuration includes initial installation, EuLLM, RAG Enterprise and 12 months of I3K AI Suite.
Ryzen AI Max+ 395 · 128 GB unified memory
from €5,900
Maximum memory capacity in the entry tier. Built for very large models, RAG over extensive corpora and document intelligence where keeping everything in memory is what matters.
32 GB dedicated GPU · Radeon AI PRO R9700 or validated equivalent
from €5,900
Same price bracket, opposite philosophy: less memory, more inference speed. Suited to MoE models, responsive RAG and agents that need to answer fast.
64 GB total · dual accelerator
from €8,900
The centre of the range. It can pool both GPUs to serve larger models, or dedicate them to separate loads — LLM on one, vision, embeddings or agents on the other — to handle concurrent workloads without them competing.
NVIDIA configuration · 32 GB
from €9,900
For teams with an explicit CUDA requirement, whether from proprietary libraries, existing pipelines or internal policy. Same software stack, NVIDIA ecosystem.
96 GB GPU with ECC memory
from €22,900
Enterprise tier. Large models, high session concurrency and ECC memory for installations where compute reliability is not negotiable.
Next-generation Ryzen AI Max+ PRO · up to 192 GB unified memory
On request
Coming soon. Built on the new unified memory generation, for workloads that need even more capacity in a single node.
Hardware, initial installation, EuLLM, RAG Enterprise and 12 months of I3K AI Suite at the agreed tier.
From the second year the subscription is optional and is chosen based on the number of licensed users.
€129/month
Up to 25 users · 2 technical contacts
€399/month
Up to 100 users · 3 technical contacts
€699/month
Up to 250 users · 5 technical contacts
€1,190/month
Up to 500 users
€1,990/month
Up to 1,000 users · maximum 10 authorised contacts
Above 1,000 users: sizing and pricing on request.
End users do not open tickets directly with us. Standard I3K support works with the administrators or technical contacts designated by the customer, up to the number allowed by the chosen tier. If you need a first-line help desk for end users, that is a separate Managed Support service we can quote independently.
The user count in each subscription tier is about licensing, not how many people can use the system at the same instant. Actual concurrency depends on the hardware: for large organisations it is sized with multiple appliances or a cluster.
The installed software does not deactivate at expiry. Without an active subscription you lose updates, security maintenance, new releases, new certified model profiles and support — but the existing installation stays operational.
Early Access means we work on sizing together before talking about orders. The first step is understanding which workloads you want to bring in-house.
Fill in your company details: we need them to size the configuration properly and put together a proposal that makes sense. We reply within two working days.