Security Whitepaper

NewMax AI is built on a Local-First architecture. This means all your data — conversations, uploaded files, app settings, model preferences — is stored entirely on your local device. Our servers don't participate in data storage and have no ability to access your local data.

• Conversations are stored in a local database — never uploaded to the cloud

• Uploaded files are processed and parsed locally; originals stay in place after processing

• App settings and user preferences are stored in local configuration files

• Uninstalling the app completely erases all data — no need to contact us for account deletion

When you use cloud AI models (e.g., OpenAI GPT, Google Gemini, Kimi, Minimax), your conversations are sent to the respective model provider for processing. Our role in this process is a transparent encrypted relay:

• Messages are transmitted via HTTPS with full TLS protection

• Zero-log relay — we don't record, cache, or store any conversation content

• No analysis, annotation, or secondary processing of message content

• Your data is never used for model training or any other commercial purposes

For scenarios with the strictest data security requirements (legal documents, medical records, financial data), NewMax AI supports local models for physical-level privacy protection:

• Supports Ollama, LM Studio, and other local model services

• Full functionality even in physically air-gapped environments

• Data is generated, processed, and stored entirely on your device

• Passes even the most rigorous enterprise security audits with confidence

Transparency is the foundation of trust. Here's the complete list of what we do and don't collect:

We encourage every user to personally verify our security commitments. Here are the methods you can use: