Fidelis Memory is a Python library that provides faithful, high-fidelity memory for AI agents and autonomous systems. It defaults to zero-LLM retrieval using integer-pointer and vector-store techniques for efficiency and accuracy. Built for developers building with frameworks like LangChain, Mem0, or custom RAG pipelines, it is available on PyPI with an MIT license.
In the AI & ML space, fidelis-memory takes a focused approach. It focuses on maintaining accurate and faithful memory for AI agents without relying on LLM calls for retrieval. fidelis-memory is an open-source project aimed at AI developers. fidelis-memory is open source under the MIT license. fidelis-memory is available on the command line, and it can be self-hosted.
Hermes Labs builds and maintains fidelis-memory, and it first shipped in 2026. The project is developed in the open on GitHub with 21 stars and 15 commits in the last 90 days. Among its 4 catalogued features are Agent Memory, Zero-LLM Retrieval, and Vector Store.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Same category — not a similarity match