What's New in Lumo 2.0

Image Generation, Recognition, and Editing

Lumo 2.0 is now a multimodal AI assistant. It can generate images from text prompts or rough sketches, and edit existing images by changing backgrounds, adjusting colors, or removing objects. It also reads and analyzes uploaded photos, charts, documents, and screenshots .

Custom Lumos

Users can create purpose-specific AI assistants with a pre-set writing style and response format. This enables tailored workflows for drafting, research, coding, or other recurring tasks .

Persistent Memory (User-Managed)

Lumo can now remember information across sessions. You control exactly what it retains: you can store, review, delete, or block specific memories. This makes conversations more coherent without sacrificing privacy .

Doubled Context Window

The context window is twice as large as before, allowing Lumo to handle longer documents and larger datasets in a single session . For very large files — even 4,000-page documents — Lumo splits them into smaller chunks, indexes them, and retrieves only the relevant parts when queried .

Thinking and Fast Modes

Users can choose between Fast mode for quick answers and Thinking mode for multi-step reasoning on complex problems .

Ghost Mode

When enabled, conversations disappear permanently as soon as you close the chat, leaving no history .

Projects (from v1.3)

Encrypted workspaces that bundle chats, files, and requirements for a task, keeping everything in sync across devices .

Redesigned UI

A brand-new design for both web and mobile apps .

Performance Improvements

Lumo 2.0 delivers a 240% performance leap over the previous version, according to Proton . The upgrade introduces two distinct model tiers:

• Max — Powered by GLM-5.2, this is the top-performing model. Proton states it surpasses Claude Sonnet 4.6 on benchmarks .
• Lite — Powered by Qwen 3.5 with 122 billion parameters, designed for faster, lighter responses .

Plus plan subscribers receive priority access to the Max model, which results in faster responses .

Privacy Protections

Privacy remains the defining differentiator for Lumo 2.0 :

• Zero-access encryption — Proton cannot read your conversations; only you can decrypt and access your chat data .
• No chat logging — Proton does not store logs of conversation history on its servers .
• No training on user data — Your data is never used to train AI models .
• Ghost mode — Ephemeral sessions with no saved history .
• End-to-end encryption — Saved chat history is protected with end-to-end encryption .
• European infrastructure — Operated under Swiss privacy laws .

The security model ensures that even Proton's systems cannot access your conversational data, using a combination of asymmetric and symmetric encryption similar to Proton Mail and Drive .

Pricing Details

Lumo 2.0 is available in three tiers. The table below summarizes the options:

Important pricing note: Multiple sources report Lumo Plus at $12.99/month (or €12.99) . One Dutch report from the v2.0 launch states $10/month , which may reflect regional pricing or a temporary promotion. Proton's own support page lists €12.99/month , and the App Store listing shows $12.99/month and €12.99/month across different regions . The annual plan is approximately $119.99 or €119.88 .

Proton Unlimited subscribers do not automatically get Lumo Plus access; it must be purchased separately . Visionary and Lifetime plan holders receive Lumo Plus at no extra cost .

Availability

Lumo 2.0 is available immediately on the web, Android, and iOS. The free tier provides a limited taste of the Max model, with the most advanced capabilities reserved for Lumo Plus subscribers .

Proton reports that Lumo has exceeded 10 million users since its initial launch in July 2025 .

What's New in Lumo 2.0

Image Generation, Recognition, and Editing

Lumo 2.0 is now a multimodal AI assistant. It can generate images from text prompts or rough sketches, and edit existing images by changing backgrounds, adjusting colors, or removing objects. It also reads and analyzes uploaded photos, charts, documents, and screenshots .

Custom Lumos

Users can create purpose-specific AI assistants with a pre-set writing style and response format. This enables tailored workflows for drafting, research, coding, or other recurring tasks .

Persistent Memory (User-Managed)

Lumo can now remember information across sessions. You control exactly what it retains: you can store, review, delete, or block specific memories. This makes conversations more coherent without sacrificing privacy .

Doubled Context Window

The context window is twice as large as before, allowing Lumo to handle longer documents and larger datasets in a single session . For very large files — even 4,000-page documents — Lumo splits them into smaller chunks, indexes them, and retrieves only the relevant parts when queried .

Thinking and Fast Modes

Users can choose between Fast mode for quick answers and Thinking mode for multi-step reasoning on complex problems .

Ghost Mode

When enabled, conversations disappear permanently as soon as you close the chat, leaving no history .

Projects (from v1.3)

Encrypted workspaces that bundle chats, files, and requirements for a task, keeping everything in sync across devices .

Redesigned UI

A brand-new design for both web and mobile apps .

Performance Improvements

Lumo 2.0 delivers a 240% performance leap over the previous version, according to Proton . The upgrade introduces two distinct model tiers:

• Max — Powered by GLM-5.2, this is the top-performing model. Proton states it surpasses Claude Sonnet 4.6 on benchmarks .
• Lite — Powered by Qwen 3.5 with 122 billion parameters, designed for faster, lighter responses .

Plus plan subscribers receive priority access to the Max model, which results in faster responses .

Privacy Protections

Privacy remains the defining differentiator for Lumo 2.0 :

• Zero-access encryption — Proton cannot read your conversations; only you can decrypt and access your chat data .
• No chat logging — Proton does not store logs of conversation history on its servers .
• No training on user data — Your data is never used to train AI models .
• Ghost mode — Ephemeral sessions with no saved history .
• End-to-end encryption — Saved chat history is protected with end-to-end encryption .
• European infrastructure — Operated under Swiss privacy laws .

The security model ensures that even Proton's systems cannot access your conversational data, using a combination of asymmetric and symmetric encryption similar to Proton Mail and Drive .

Pricing Details

Lumo 2.0 is available in three tiers. The table below summarizes the options:

Important pricing note: Multiple sources report Lumo Plus at $12.99/month (or €12.99) . One Dutch report from the v2.0 launch states $10/month , which may reflect regional pricing or a temporary promotion. Proton's own support page lists €12.99/month , and the App Store listing shows $12.99/month and €12.99/month across different regions . The annual plan is approximately $119.99 or €119.88 .

Proton Unlimited subscribers do not automatically get Lumo Plus access; it must be purchased separately . Visionary and Lifetime plan holders receive Lumo Plus at no extra cost .

Availability

Lumo 2.0 is available immediately on the web, Android, and iOS. The free tier provides a limited taste of the Max model, with the most advanced capabilities reserved for Lumo Plus subscribers .

Proton reports that Lumo has exceeded 10 million users since its initial launch in July 2025 .