Proton has launched Lumo 2.0, the first major upgrade to its privacy‑focused chatbot since debuting a year ago. The new version adds image recognition and generation, letting users create, edit, and analyze pictures within a conversation. All image‑related chats remain protected by zero‑access encryption, meaning only the user’s device can read the data.

Lumo 2.0 also introduces a “thinking mode” for deeper reasoning and can pull up background information with source citations. In Proton’s internal Artificial Analysis Intelligence Index benchmark, the Lite model posted a 127 % improvement over version 1.4, while the Max configuration jumped 240 %. Those gains narrow the gap with OpenAI’s latest models.

Core chat functions stay free, but unlimited conversations, advanced image generation, and access to the higher‑tier models require a $10 monthly Lumo Plus subscription. By bundling strong privacy guarantees with capabilities that rival mainstream AI services, Proton positions Lumo as a viable choice for security‑conscious users and enterprises wary of data harvesting.

Analysts note that Lumo’s privacy‑first architecture could pressure larger AI firms to offer on‑device encryption options. If adoption grows, the subscription model may inspire other niche players to monetize premium features without compromising user data. For now, Proton’s move shows that privacy and advanced AI can coexist in a single consumer product.