AI agent memory / LLM memory
MemPalace is a Python AI-memory library that stores conversation history verbatim and retrieves it via scoped semantic search using a pluggable backend.
Same problem, same approach. Swapping one for another is a config change, not a rewrite.
Use when you want an actively maintained universal memory layer with interoperable storage and retrieval.
Universal memory layer for AI Agents. It provides scalable, extensible, and interoperable memory storage and retrieval to streamline AI agent state management for next-generation autonomous systems.
Use when you want a maintained memory library and can accept AGPL-3.0 instead of MIT.
Memory library for building stateful agents
Solves the same problem with a different architecture or at a different layer. Expect to rewrite the integration.
Use when you want a learning memory model that adapts over time instead of verbatim storage.
Hindsight: Agent Memory That Learns
Use when you want a Rust single-file memory layer with no separate database instead of a Python library with pluggable backends.
Memory layer for AI Agents. Replace complex RAG pipelines with a serverless, single-file memory layer. Give your agents instant retrieval and long-term memory.
Use when you need self-evolving memory with skill reuse and token savings instead of verbatim search.
Self-evolving memory OS for LLM & AI Agents: ultra-persistent memory, hybrid-retrieval, and cross-task skill reuse, with 35.24% token savings and DeepSeek Harness support.
Use when you want a standalone context retrieval layer rather than embedding a Python memory library.
Open-source context retrieval layer for AI agents
Use when you want research-grade agentic memory that builds LLM-generated notes instead of verbatim records.
A-MEM: Agentic Memory for LLM Agents
An earlier or more limited way to do the job. Still the right call when you need something small, proven, or CPU-only.
Use only if you accept a stale, unmaintained local-memory module; last push is over 1.7 years ago.
Your Local Artificial Memory on your Device.
Does the same job, but ships as an app. Useful to a person, not swappable into a codebase.
Use when you want a ready-made CLI/plugin to capture agent sessions rather than embedding a memory library.
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
These projects were analysed and named mempalace among their alternatives. The relationship is not symmetric — how mempalace rates them is a separate judgement, made when mempalace is analysed in its own right.
calls mempalace “Choose when you want a MIT-licensed memory system with strong benchmark claims.”
Universal memory layer for AI Agents
calls mempalace “Use it if you want a Python/MIT self-hosted memory system that claims top benchmark scores.”
Memory and context engine + app that is extremely fast, scalable, and can be run fully locally. The Memory API for the AI era.
calls mempalace “if you want an MIT-licensed memory system with published benchmarks and no vendor lock-in.”
TencentDB Agent Memory is a team-level memory hub for AI Agents — turning conversations, docs, and code into four reusable memory assets (Chat Memory, Skill, LLM-Wiki, Code-Graph) that are governed, shared, and equipped across agents and frameworks.
calls mempalace “Choose if you want an MIT-licensed memory system with claimed benchmark leadership and permissive licensing.”
Memori is agent-native memory infrastructure. A LLM-agnostic layer that turns agent execution and conversation into structured, persistent state for production systems. Built for enterprise, Memori works with the data infrastructure you already run, no rip-and-replace, and deploys across managed cloud, single-tenant cloud, VPC, and on-premises.
calls mempalace “Python-native memory layer with strong benchmarks as a drop-in alternative.”
Self-evolving memory OS for LLM & AI Agents: ultra-persistent memory, hybrid-retrieval, and cross-task skill reuse, with 35.24% token savings and DeepSeek Harness support.
calls mempalace “Choose when you prefer an MIT-licensed, benchmark-focused memory system over Honcho's AGPL service.”
Memory library for building stateful agents
calls mempalace “choose for an MIT-licensed, benchmark-optimized memory system if you favor permissiveness.”
Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.
calls mempalace “Choose it if you want a benchmark-optimized, MIT-licensed memory system with a lighter footprint than cognee's graph engine.”
Cognee is the open-source AI memory platform for agents. Give your AI agents persistent long-term memory across sessions with a self-hosted knowledge graph engine.
calls mempalace “Use if you want a Python, MIT-licensed memory system with a benchmark focus rather than a drop-in TypeScript module.”
#1 Persistent memory for AI coding agents based on real-world benchmarks
calls mempalace “Choose if you want an MIT-licensed Python memory system with its own API rather than Hindsight's server/client model.”
Hindsight: Agent Memory That Learns
calls mempalace “choose it when you want a benchmark-focused memory layer with an alternative storage/retrieval design.”
One portable memory layer for every AI agent: local-first, Markdown-native, user-owned, and self-evolving across apps, tools, and workflows.
calls mempalace “Choose it when you want an agent memory system instead of a multi-source retrieval and search API.”
Open-source context retrieval layer for AI agents