Private models
Connect to a private model inside the practice’s network.
Deployment
A separate environment for each practice.
Choose where Cortega Legal runs and which models it uses. Each practice has its own web address and standalone infrastructure.
Choose where it runs.
On premises, in the practice’s AWS account or on rented GPUs. Practice data is not shared with other practices.
Discuss network access, model providers and the documents your team works with when booking a demo.
Foundation models and provider keys
Use foundation models, private models together with foundation models, or the practice’s own provider keys.
Practice address
Application sign-in runs at the practice’s own Cortega Legal address.
Deployment details
Can the application run inside our network?
Yes. On-premises deployment and a deployment in your own AWS account are available.
Can we combine private and foundation models?
Yes. A practice can connect private models, foundation models or both, using its own provider keys.
How do model choices affect cost and network control?
Open-weight models can run on infrastructure inside your network or VPC, subject to their licence and hardware requirements. Proprietary models accessed through external providers require those providers’ endpoints and data terms. A more capable model may improve some tasks, but quality and cost must be measured in your workflow.
Can a provider offer an air-gapped deployment?
Assess the provider’s actual network boundary and all dependencies. A fully air-gapped configuration needs local model inference, research data and verification services; external research connectors require network access. Private hosting and a provider’s data-retention policy are separate choices.
Discuss deployment for your practice.
Review the environment, model choices and document workflow.