Neura Market
**Issues** logo

**Issues**

Free

Open-source persistent memory layer for AI agents and applications

FreeFree tier
Type
Open Source
Visit Website GitHub (40 stars)

About **Issues**

Memori is an open-source AI memory management layer designed to provide persistent, structured memory for AI agents and applications. The MemoriLabs/Memori repository hosts the core codebase and tracks development via GitHub Issues, where users request features such as pluggable retrieval index backends, Milvus integration, native CrewAI Framework support, batch memory extraction, and local custom OpenAI-compatible endpoint support (Ollama/vLLM). With over 15,000 stars and active community contributions, Memori aims to enable long-term memory, context retention, and cross-session persistence for conversational AI and automation workflows.

Key Features

Pluggable retrieval index backend with optional Milvus support
Native integration with CrewAI Framework via Python SDK
Support for local custom OpenAI-compatible endpoints (Ollama/vLLM)
Batch and on-demand memory extraction instead of per-message processing
Native-language fact extraction for multi-language support
Project-scoped BYODB conversation identity and trace metadata
Memory deletion capabilities for user privacy management
Neon Launchpad DB provisioning for scalable storage

Pros & Cons

Pros
  • Open-source and free to use with active community contributions
  • Highly extensible via pluggable backends and SDK integrations
  • Supports both cloud and local deployment for privacy-sensitive use cases
  • Offers batch extraction capabilities to reduce per-message overhead
  • Actively developed with frequent feature requests and bug fixes
  • Large and growing community (15.6k stars, 2.9k forks)
Cons
  • Documentation may lag behind feature development (as seen in issue #614)
  • Some requested features (e.g., Dark Mode Toggle) are minor UI enhancements, not core memory functionality
  • Dependency on GitHub for issue tracking may fragment user support
  • Lack of native bundling into certain frameworks (e.g., Hermes) requires manual integration

Best For

Enhancing conversational AI with long-term memory and context persistenceBuilding AI agents that recall past interactions across sessionsIntegrating memory into multi-agent frameworks like CrewAIDeploying local, offline memory systems using Ollama/vLLM endpointsManaging user-specific memories in applications requiring data deletionSupporting multilingual fact extraction for global deployments

FAQ

Can I delete user memories in Memori?
Yes, there is an open issue (#303) discussing how to delete user memories, indicating the feature is being addressed.
Does Memori support local AI endpoints like Ollama or vLLM?
Yes, a feature request (#250) asks for support of local custom OpenAI-compatible endpoints (Ollama/vLLM), showing community interest and planned support.
Can I integrate Memori with CrewAI?
Yes, there is a feature request (#495) for CrewAI Framework Integration for the Memori Python SDK, indicating planned compatibility.