NCM Memory

NCM Memory is a locally running software layer designed to provide persistent, user-controlled memory for language models and AI assistants. Inspired by neurobiology, it introduces a unified, model-agnostic memory system that enables continuity of conversation and long-term context across multiple supported AI models, including ChatGPT, Claude, Gemini, Perplexity, and Ollama. The core technology was originally developed for integration with BioCortexAI and can be linked to various large language models through a wrapper, allowing users to maintain their conversational context regardless of the underlying AI model in use.

A key feature of NCM Memory is its commitment to privacy and user ownership. All data is stored and encrypted locally on the user's own device, with encryption tied to hardware signatures. The software provider does not have access to encryption keys or the ability to read memory content. Users have full control over when memory is used, what information it contains, and when it is deleted or modified. Memory content is only shared with AI models as system context after explicit user approval, ensuring transparency in what information is provided to the model. ncm file, facilitating easy transfer of context between models or users, and enabling the concept of "Practice-as-a-Product," where accumulated experience and decision history can be handed over to teams or clients.

NCM Memory offers a web plugin that integrates with the native web interfaces of ChatGPT, Claude, and Gemini, allowing the memory layer to be used directly within these environments. The software is available for Windows and Linux platforms. Performance is optimized to handle years of conversational data without significant slowdowns, employing a dynamic structure of relationships, experiences, and meanings rather than a traditional database approach. The tool reduces repetitive prompts and can decrease token usage by up to 80% in initial queries.

Pricing includes a free "Chat & Creatives - Trial" plan with 30 days of access to the web extension and local Ollama integration, as well as a paid "Chat & Creatives" plan at a discounted rate for early adopters, offering unlimited web extension use, full local access, and export capabilities. Additional plans for coding and enterprise use are mentioned as forthcoming. NCM Memory is positioned as a privacy-focused, persistent memory solution for individuals and teams working with language models and AI assistants.