SaaSHub helps you find the best software and product alternatives Learn more →
Top 23 Python ai-memory Projects
-
cognee
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.
Project mention: I Merged 348 PRs. Here Are The 11 Bugs That Tested My Sanity. | dev.to | 2026-07-15 -
SaaSHub
SaaSHub - Software Alternatives and Reviews. SaaSHub helps you find the best software and product alternatives
-
julep
Julep — durable, composable AI agents. Flows that crash and resume, retry safely, and explain every step.
Project mention: Show HN: BaaS to build agents as data, not code | news.ycombinator.com | 2025-08-08 -
Pinned Honcho source tree
-
-
iai-personal-memory-engine
Independent Autistic Intelligence — a cyber brain for your AI. It never forgets a detail, remembers exactly what you said, and learns how you work over time. Free, local, works with Cursor, Claude Code, Codex, OpenClaw, Hermes and more. MIT.
Project mention: 50K-token context reloads, replaced by a 2K local memory engine | news.ycombinator.com | 2026-06-28 -
pmb
Local-first persistent memory for AI coding agents (Claude Code, Cursor, Codex) over MCP. Decisions, lessons and facts live in one SQLite file on your disk. Offline, multilingual.
Project mention: Show HN: PMB – local memory for coding agents that shows if it is used | news.ycombinator.com | 2026-07-01Repo for anyone who wants to look under the hood: https://github.com/oleksiijko/pmb
-
empirica
Make AI agents and AI workflows measurably reliable. Epistemic measurement, Noetic RAG, Sentinel gating, and grounded calibration for Claude Code and beyond
Project mention: Why Your AI Agent Needs Memory That Decays (and How Qdrant Makes It Work) | dev.to | 2026-03-06GitHub: github.com/Nubaeon/empirica
-
superlocalmemory
World's first local-only AI memory to break 74% retrieval and 60% zero-LLM on LoCoMo. No cloud, no APIs, no data leaves your machine. Additionally, mode C (LLM/Cloud) - 87.7% LoCoMo. Research-backed. arXiv: 2603.14588
Memory: SLM-Lite cognitive memory, powered by SuperLocalMemory — backed by 3 peer-reviewed papers (arXiv:2604.04514, arXiv:2603.14588, arXiv:2603.02240). SQLite-backed, fully local. Episodic memory, semantic recall via embeddings (new in v2.2.0), working memory with decay. No cloud dependency. Your agent's memory stays on your machine.
-
mengram
Human-like memory for AI agents — semantic, episodic & procedural. Experience-driven procedures that learn from failures. Free API, Python & JS SDKs, LangChain, CrewAI & OpenClaw integrations.
Website: mengram.io
-
piia-engram
Local-first AI memory you can see, edit, and override — portable across Claude Code, Codex, Cursor, Windsurf, and other MCP coding tools.
I packed this into piia-engram — a local-first, content-blind memory layer for MCP clients (Claude Code, Cursor, Codex). Open source, pip install piia-engram. It's local-first mostly because the cleanest answer to "is my memory private?" is "nothing left your machine in the first place."
-
Project mention: Alive – Five Markdown files that give Claude Code a persistent memory | news.ycombinator.com | 2026-03-17
- insights.md — stuff that doesn't change
Save at session end, load at session start. 12 hooks enforce the structure, log entries are immutable, deletions are blocked, context re-injects when the window fills up. The model can't talk its way out of the guardrails.
Also works with OpenClaw (separate plugin, same files - in alpha). Your Mac Mini captures all day, Claude Code does focused work, both read the same project context.
claude plugin install alive
MIT licensed, local files, no account.
https://github.com/alivecomputer/alive-claude
-
vektori
Memory that remembers the story not just the facts. Three layer sentence graph for AI agents -> Facts, Episodes, raw Sentences. One DB. Zero config.
My own investigation of the memory research last week kept landing on a related but distinct conclusion. Three-tier layered memory (fast index on top, episodes in the middle, raw transcripts on demand) is now the default pattern in serious memory systems. LongMemEval-S at 73% is a reasonable community-stack baseline (Vektori, BGE-M3 plus Flash-2.5-lite), and MemPalace's ChromaDB-backed stack has since posted 96.6% on LongMemEval R@5 in raw mode, which is the kind of jump worth watching. The leaked Claude Code memory internals, three subsystems plus a "Dream" consolidation pass, is almost exactly this architecture. The failures in agent memory aren't mostly retrieval: they sit in the reasoning layer between retrieval and action, and in the moment a session ends and all the volatile context evaporates.
-
mem0-mcp-selfhosted
Self-hosted mem0 MCP server for Claude Code. Run a complete memory server against self-hosted Qdrant + Neo4j + Ollama while using Claude as the main LLM.
Project mention: Show HN: Giving Claude Code persistent memory with a self-hosted MCP server | news.ycombinator.com | 2026-02-19 -
Project mention: Show HN: Memory System for Claude Code and Other CLIs | news.ycombinator.com | 2025-12-05
-
-
TrueMemory
A living memory system that ingests long-horizon data to infer insights, enabling more decisive action, all while running on a single SQLite file locally.
Project mention: I tried to preserve my grandmother's mind. It became a dev tool | news.ycombinator.com | 2026-05-21 -
widemem-ai
Next-gen AI memory layer with importance scoring, temporal decay, hierarchical memory, and YMYL prioritization
Project mention: Your AI Memory System Can't Tell a River Bank from a Savings Account | dev.to | 2026-04-14The full change is in PR #16. Eight files changed, 163 tests passing, zero additional API calls.
-
sibyl
Collective intelligence runtime for AI agents. Knowledge graph + persistent memory. (by hyperb1iss)
Project mention: Show HN: Sibyl – self-hosted cross-agent memory for AI coding agents | news.ycombinator.com | 2026-06-30 -
Octopoda-OS
The open-source memory operating system for AI agents. Persistent memory, semantic search, loop detection, agent messaging, crash recovery, and real-time observability.
Project mention: Engineer made built in memory, loop detection and audits mandatory for agents | news.ycombinator.com | 2026-07-16 -
agents-remember
The system of record and control plane for AI coding agents. Keeps their memory correct, current, and safe to act on as the code moves, captures what code can't say on its own, and gates what agents are allowed to do. Retrieves by path, semantic search, and relationship (code-graph).
Project mention: Show HN: Agents Remember – Git-aware memory for coding agents | news.ycombinator.com | 2026-06-05 -
brainapi2
BrainAPI is a knowledge graph–powered AI memory layer that transforms unstructured data into structured knowledge, enabling intelligent search, recommendations, and contextual memory for AI agents and applications.
Project mention: Show HN: BrainAPI's event-centric graph from any data with four-stage pipeline | news.ycombinator.com | 2026-06-25 -
Project mention: The trick to AI coding memory isn't a bigger instruction file — it's smaller, layered knoledge | dev.to | 2026-04-06
I wrote up the full system with templates and an automated setup script and put it on GitHub if anyone wants the details: https://github.com/sms021/SuperContext
-
Context-Packer
A Python script that bundles your project's source code into a single text file. Perfect for providing full context to AI assistants like ChatGPT, Gemini, and Claude.
NOTE:
The open source projects on this list are ordered by number of github stars.
The number of mentions indicates repo mentiontions in the last 12 Months or
since we started tracking (Dec 2020).
Python ai-memory discussion
Python ai-memory related posts
-
Engineer made built in memory, loop detection and audits mandatory for agents
-
I beat mem0 on long eval memory and could not care less
-
Show HN: PMB – local memory for coding agents that shows if it is used
-
Show HN: BrainAPI's event-centric graph from any data with four-stage pipeline
-
In April, a Claude built a tool to leave notes for future Claudes. In June, I showed up.
-
Create Agents that are Traceable, detect loops and have a shared memory
-
I Tested 33 AI Memory Engines — Here's What Actually Works
-
A note from our sponsor - SaaSHub
www.saashub.com | 17 Jul 2026
Index
What are some of the best open-source ai-memory projects in Python? This list will help you:
| # | Project | Stars |
|---|---|---|
| 1 | cognee | 27,949 |
| 2 | julep | 6,597 |
| 3 | honcho | 5,910 |
| 4 | memobase | 2,776 |
| 5 | iai-personal-memory-engine | 340 |
| 6 | pmb | 305 |
| 7 | empirica | 237 |
| 8 | superlocalmemory | 197 |
| 9 | mengram | 183 |
| 10 | piia-engram | 170 |
| 11 | alive | 118 |
| 12 | vektori | 112 |
| 13 | mem0-mcp-selfhosted | 100 |
| 14 | memory | 91 |
| 15 | memv | 86 |
| 16 | TrueMemory | 84 |
| 17 | widemem-ai | 47 |
| 18 | sibyl | 44 |
| 19 | Octopoda-OS | 30 |
| 20 | agents-remember | 26 |
| 21 | brainapi2 | 25 |
| 22 | RunawayContext | 16 |
| 23 | Context-Packer | 13 |
Sponsored
SaaSHub - Software Alternatives and Reviews
SaaSHub helps you find the best software and product alternatives
www.saashub.com