⬢github Python · 1.1K ★ +182 since we first saw it · pushed 20 min ago · NOASSERTION
Tencent-Hunyuan/Hy-Memory
Hy Memory is Tencent Hunyuan's open-source memory engine for LLM agents, offering specialized memory pipelines for chat (long-term cognitive memory), working (facts and reusable experience), and coding (version-evolution diff tracking). It ships as a Python SDK plus MemHub, a local desktop/CLI tool that unifies conversation history and coding traces from tools like Claude Code and Cursor into one managed local memory runtime.
Why now: Fresh October 2026 open-source release: Hy Memory 2.0 on PyPI, the Hy-Memory-30B-A3B model with a technical report, and MemHub v1.0.4 with signed desktop builds.
Who it is for: Developers building self-evolving LLM agents or AI coding tools who want persistent, governed cross-application memory.
Stars over our 210 snapshots: 958 to 1.1K, since 2 d ago.
Where people talked about it
- ⬢github new repos, most starred 8 min ago
- ⬢github new repos, most starred 8 h ago
- ⬢github new repos, most starred 1 d ago
API: https://socialmediatrends-api.osmike.com/v1/repos/Tencent-Hunyuan/Hy-Memory