LLM series

AI for mission-critical use cases

Large Language Models built with advanced technology, tailored for regulated industries.

Colosseum 355B
A cutting-edge
multi-language model
for Agentic AI use cases
Italia 10B
A lightweight
multi-language model for entreprise use cases

Benchmarks

State-of-the-art performance

This is an analysis of Colosseum 355B’s performance against other comparable LLMs.

Value proposition

Why choose iGenius LLMs

Leverage our private models to get accurate, contextualized answers in a safe and compliant way.

For regulated industries
iGenius LLMs were built for industries with strict compliance standards, such as financial services, government and heavy industry.
Mission-critical use cases
Easy to fine-tune to specific business data, iGenius LLMs are ideal for mission-critical use cases that require personalized AI models.
FP8 Format
Trained in the FP8 format, iGenius’ language models optimize the use of computational resources and accelerate calculation operations, reducing costs and energy consumption without compromising the models’ performance.
Continual Pre-Training
Continual Pre-training with proprietary company data allows an LLM to incorporate domain-specific knowledge and adapt to your organization’s requirements, with continuous updates and refinements to deliver the most accurate, customized results every single time.
Privacy and Security
The architecture of iGenius’ language models is designed to adhere to elevated security and compliance standards in regulated sectors.


Cost efficient
Sustainable and highly efficient, iGenius LLMs enable business growth with optimal financial and resource investment.

Business application

How to leverage iGenius LLMs in your organization
Continual Pre-training
Continual Pre-training (CPT) is a key technique through which your company can transform a general-purpose Large Language Model (LLM) into a truly proprietary asset. By continuing the training process with proprietary company data, CPT incorporate domain-specific knowledge into the model, adapting it to the unique requirements of your organization.
Fine-tuning
Our AI models allow you to tailor the training data to reflect your specific domain, fine-tuning it to increase accuracy and ensure responses in line with your business needs.
Line chart
Retrieval-augmented generation
By connecting to your company’s data sources and retrieving up-to-date, verified information, our models provide accurate and relevant responses to user queries.
it