Skip to main content
OpenSourceAlternatives

Proton Lumo

Open source · OSI-approvedStatus: unknownNew

Private AI chat with zero-access encrypted history

Alternative to ChatGPT

Proton Lumo is a privacy-first AI chat assistant from the Swiss company behind Proton Mail and Proton VPN. It answers questions, analyzes documents, searches the web, and generates images while storing every conversation with zero-access encryption, keeping no logs, and never using chats to train models. It serves people who want everyday AI help without feeding a surveillance business model.[1][2]

Quick verdict

Privacy-motivated individuals and small teams handling sensitive material benefit most from Lumo. Its concrete strength is confidential everyday assistance: encrypted chat history that even Proton cannot read, paired with image generation, web search, and document analysis in one assistant. The material constraint is capability depth and usage caps: the free plan limits model usage, messages, history, and image generations, and power users may outgrow its boundaries. If you want AI help that cannot profile you, choose Lumo; if you need the deepest reasoning, largest context windows, or API access, choose a mainstream assistant and accept its data terms.[1][2][3]

Key features

  • Zero-access encrypted history Saved conversations are encrypted so only you can read them; neither Proton nor any third party can access their contents.[1][2]

  • No-log, no-train posture Lumo keeps no chat logs and never uses conversations to train AI models, build profiles, or serve targeted advertising.[1][2]

  • Account-optional start Anyone can begin chatting for free without creating a Proton account, and signing in unlocks encrypted history synchronized across devices.[1][2]

  • Image generation Text descriptions become images in seconds with iterative refinement, and personal designs are never reused to inspire other outputs.[1][2]

  • Real-time web search Built-in search answers questions about current events without tying queries to an advertising profile.[1][2]

  • Confidential document analysis Uploaded reports and research can be summarized and analyzed without their contents entering a training set.[1]

  • Projects and Custom Lumos Extended work can be organized into Projects with specialized Custom Lumos assistants, with unlimited allowances on the Plus plan.[1][2]

  • Open-source clients The web and mobile clients are openly licensed under GPL-3.0 so anyone can verify the privacy claims in code, while the model-serving backend stays proprietary.[1][2][3]

Who will love it

  • Surveillance-weary everyday users People who ask AI about health, work, or ideas get useful answers with a structural guarantee that questions cannot become ad profiles.[1][2]

  • Confidential-document teams Staff summarizing reports or research under GDPR get analysis without contents entering a training set, backed by a European service with account deletion.[1][2]

  • Proton households Existing Proton users extend the same encrypted-vault logic they trust for mail and drive to AI chats, synced across web, iOS, and Android.[1]

Who should avoid it

  • Frontier-capability chasers Users who need the strongest reasoning, largest context windows, or cutting-edge model features should verify Lumo's current models against mainstream flagships before committing.[1]

  • Heavy free-tier users The free plan caps Max model usage, messages, history, image generations, Projects, and Custom Lumos, so daily intensive users will hit limits without Plus.[1]

  • Builders needing an API No public API is documented for Lumo, so developers wanting programmatic access or custom integrations should choose a platform that offers one.[1]

Deployment and fit

European hosted service Lumo runs as a Proton-hosted European service under GDPR with account deletion; the free plan starts without an account and Plus or Business unlock higher limits.[1]

Web, iOS, and Android reach The assistant is reachable through the lumo.proton.me web app and official iOS and Android apps, with encrypted history sync for signed-in accounts.[1]

Questions worth asking

What is the best Proton Lumo alternative for non-technical users?

If Lumo's free-plan limits frustrate you, ChatGPT is the conditional choice: it offers broader capability and generous free access for casual users, with the evidence-backed tradeoff that your conversations are processed on OpenAI infrastructure rather than held under zero-access encryption.[1][2]

Is Proton Lumo really no-code?

Day-to-day use is visual and no-code: chatting, uploading files, generating images, and organizing Projects all happen through the chat interface. The open-source clients matter only to reviewers reading the code; running Lumo needs no development work, though no public API exists for builders.[1][2]

How much does Proton Lumo cost per month?

As checked October 12, 2026, Lumo is free at $0.00 with limited usage, while Lumo AI Plus costs $12.99 per month on monthly billing or $9.99 per month billed annually at $119.88 USD, with a 30-day money-back guarantee and a separate Lumo for Business plan. Recheck the official pricing page before deciding, as plans change.[1]

Can you self-host Proton Lumo?

No. The web and mobile clients are open source under GPL-3.0, but the model-serving backend is proprietary and no self-hosted server offering is documented. A repository alone does not prove self-hosting support, so treat Lumo as a European-hosted service.[1][2]

Sources and editorial review

Source research. Reviewed on October 12, 2026. Written by Sanjib Sah. AI assisted; reviewed by a human editor.

We have not run or tested this product.

  1. About Lumo, the privacy-first AI assistant

    official · Observed October 12, 2026

    Official Lumo product page: zero-access encryption, open-source code, no training, GDPR service, web/iOS/Android apps, features and plans.

  2. Lumo pricing and plans

    official · Observed October 12, 2026

    Official Lumo pricing: Free $0.00 limits, Plus $12.99 monthly or $9.99 annual, money-back guarantee, Business plan, account-optional FAQ.

  3. Introducing Lumo: AI where every conversation is confidential

    official · Observed October 12, 2026

    Official launch announcement: no logs, zero-access encrypted history, free no-account start, European hosting outside US jurisdiction.

  4. WebClients applications/lumo directory

    official · Observed October 12, 2026

    Canonical GitHub repository ProtonMail/WebClients: applications/lumo directory verified on main; hosts the Lumo web client.

  5. ProtonLumo ios-lumo repository

    official · Observed October 12, 2026

    Official ProtonLumo iOS client repository with GPL license terms; org also hosts android-lumo.

Compared with these products

Similar projects

Related projects with shared categories, technologies, or licensing.

Ollama

OllamaAIOpen

Run open language models locally with one command

Stars179.9K
Commit14h ago
LicenseMIT

Ollama reduced local model hosting from an afternoon of CUDA troubleshooting to a single pull command. It handles model download, quantisation selection, GPU offloading, and memory management, and exposes an OpenAI-compatible endpoint on localhost.

Open WebUI

Open WebUIAIOpen

One chat interface for every model you run or connect

Stars150.5K
Commit10h ago
LicenseBSD-3-Clause

Open WebUI is the front end that turns a local inference server into something a household or a team can actually use. It handles accounts, conversation history, document retrieval, and model switching, and speaks to both local runtimes and hosted APIs.

GPT4All

GPT4AllAIOpen

Private desktop chatbot that runs open LLMs fully on-device

Stars77.4K
Commit1y ago
LicenseMIT

GPT4All is a free desktop app from Nomic that runs open language models entirely on your own computer, with no cloud calls.

AnythingLLM

AnythingLLMAIOpen

Self-hosted AI workspace for chatting with documents and running agents

Stars66.7K
Commitjust now
LicenseMIT

AnythingLLM is a self-hostable AI workspace for chatting with documents, running AI agents, and switching between local and hosted language models.

Cherry Studio

Cherry StudioAIOpen

Desktop AI chat client for cloud and local models in one place

Stars52.3K
Commitjust now
LicenseAGPL-3.0

Cherry Studio is a desktop chat client for cloud and local models in one interface. It runs on Windows, macOS, and Linux, and stores data on the local machine.

LocalAI

LocalAIAIOpen

Self-hosted OpenAI-compatible runtime for local AI models

Stars49.5K
Commitjust now
LicenseMIT

LocalAI is a self-hosted runtime that runs AI models locally behind OpenAI-compatible endpoints.