Proton has unveiled Lumo 2.0, a major upgrade to its zero-access encrypted AI assistant. Built on a new architecture, the release brings the assistant closer to frontier AI models with new AI models, multimodal capabilities, Memory, improved web search, and enterprise features.
The free plan includes Lumo’s core features. Lumo Plus subscribers get unlimited chats, Projects, advanced image generation, and priority access to the fastest models. Lumo Professional is aimed at organizations, adding secure collaboration features and centralized team management.
“Lumo 2.0 is our biggest leap in capability yet. On the Artificial Analysis Intelligence Index, an independent benchmark combining 10 evaluations across agents, coding, scientific reasoning, and general knowledge, Lumo 2.0 Lite scores 127% higher than Lumo 1.4, and Lumo 2.0 Max scores 240% higher,” Eamonn Maguire, Director of Engineering, ML & AI at Proton, said.
Model types and reasoning modes
Lumo 2.0 Lite handles general questions, writing assistance, summarization, and other everyday tasks. Lumo 2.0 Max provides stronger reasoning, broader contextual understanding, and more precise outputs, making it better suited for complex or demanding tasks.
Fast mode is designed for straightforward queries, quick lookups, and tasks that benefit from rapid responses. Thinking mode generates more detailed answers for problems requiring deeper, multi-step analysis.
According to Proton, everyday queries are processed up to 76% faster than Lumo 1.4, while complex requests include a visible thinking state that shows the reasoning process.
Memory, Projects, and Custom Lumos
User-controlled Memory and Projects support ongoing collaboration across multiple conversations.
Memory retains user preferences, work styles, and conversation context across chats. Users decide what the assistant remembers, deletes, or never stores.
The context window is now twice as large, enabling the assistant to handle longer conversations and process larger documents and datasets. Projects keep related chats, files, and instructions together in encrypted workspaces across multiple sessions.
Custom Lumos let users create AI assistants tailored to specific tasks. Each can have its own instructions, writing style, and reference files, ensuring consistent responses without repeated prompts. Custom assistants can be edited, duplicated, or deleted at any time.
Multimodal capabilities
The assistant can analyze uploaded images, documents, charts, and screenshots, edit existing images, generate visuals from text prompts or sketches, and create new images within the same conversation. Web search has also been upgraded with live information and source citations, covering recent news, financial data, and weather.
Enterprise offering
Lumo for Business includes centralized administration and privacy-focused safeguards for organizations. The company says conversations are protected with zero-access encryption, are never logged or used to train AI models, and remain on the company’s European infrastructure. Administrative controls allow organizations to manage team access.
