Nuramem provides a persistent, cross-model memory layer designed to work across multiple AI assistants and platforms. It addresses the challenge that conversational AI models, such as Claude, ChatGPT, and Gemini, typically do not retain user context or decisions between sessions, and each model maintains its own isolated memory. Nuramem acts as an intermediary, creating a single, typed, audit-trailed record of user decisions, plans, references, and current state, which can be saved and recalled instantly across different models and platforms.
The tool is not a simple chat history or note-taking solution. Instead, it maintains a contract-driven, auditable memory that users can interact with by naming "Nura" in their instructions. This ensures that information is stored in Nuramem's cross-model memory, rather than in the assistant's local memory. Users can save decisions, deadlines, and project updates, then retrieve this information from any supported AI assistant in subsequent sessions. Nuramem also captures persistent information automatically and allows manual entries when needed.
Nuramem integrates with various platforms through the MCP (Model Connector Protocol), supporting clients such as Claude Desktop, Claude Code, Cursor, ChatGPT (with Plus or Business/Enterprise), and Gemini CLI. It also offers a REST API for programmatic access, enabling integration with custom agents or applications. The tool manages memory in its own managed cloud storage, with data encrypted and isolated per user. Users can export their memory at any time and are assured that deletion is permanent.
For teams, Nuramem supports shared project memory, allowing multiple users to access and contribute to a common memory layer while retaining their private personal memories. When conflicting decisions are recorded, the tool surfaces both perspectives, including who made each decision, to facilitate resolution. Team members can be invited via the management app or through supported assistants. The management app also allows browsing, searching, exporting memory, and managing shared projects and API access.
Currently, Nuramem is free for all users. Setup involves adding the Nuramem connector in the desired assistant, signing in via OAuth, and then using the designated commands to store and retrieve cross-model memory.
In the Other AI space, nuramem takes a focused approach. Enabling developers to capture and recall AI memory across models directly from the terminal. It is built as a B2B product for developers. nuramem is a paid product. It runs on the web and the command line.
nuramem first shipped in 2026. Key capabilities include memory capture, memory recall, and terminal integration. It exposes integrations via an MCP server.
Latest indexed changes and source events
Other apps tracked under the same category.
