Caffeine.ai upgrade adds model selection and privacy-focused deployment options

A new upgrade to Caffeine.ai gives developers more control over how they build and deploy applications on the Internet Computer Protocol (ICP), including the ability to select AI models and choose the type of subnet hosting their apps.
Dominic Williams, founder and chief scientist of the DFINITY Foundation, announced the update, encouraging users to try building an application to experience the changes.
One of the new options allows developers to choose the AI model they want to use. They can also select an ICP subnet based on the features their application requires, including a confidential, privacy-preserving subnet using AMD SEV-SNP technology.
The update also introduces greater flexibility in deployment choices. AMD SEV-SNP, or Secure Encrypted Virtualisation-Secure Nested Paging, is a hardware-based technology designed to protect data in use by encrypting a virtual machine’s memory and helping isolate it from the underlying host system. It is used in trusted execution environments (TEEs).
Caffeine describes its platform as an AI-powered app builder that allows users to create applications through conversations with AI. Its documentation also outlines options for selecting different AI models, which can vary in capability and cost.
The latest announcement gives developers more choice over the AI tools and infrastructure behind their applications. However, the announcement does not provide performance figures or independent security assessments for the new options, so the practical impact of the upgrade remains to be seen.
Users can explore the platform at caffeine.ai.