88Z AI, 19 October — Tencent Hunyuan has released Hy Memory 2.0, an open-source agent memory framework featuring specialized models, the framework itself, and the MemHub application. Covering chat, work, and code memories, it ranks first among open-source solutions in four benchmarks. The core Hy-Memory-30A3B model has approximately 3 billion activated parameters and outperforms Opus5 in average scores. MemHub connects with ten AI tools like Cursor for real-time memory synchronization and is already deployed in Tenc
At a glance
- Hy Memory 2.0 open-source model tops four benchmarks, outperforming Opus5
- MemHub supports memory sync across ten AI tools and is deployed in Tencent services
The update
Tencent Hunyuan has officially launched Hy Memory 2.0, an agent memory framework that open-sources specialized models, the memory framework, and the MemHub application. Focusing on chat, work, and code memory scenarios, the framework achieved first place among open-source solutions in four authoritative benchmarks. The Hy-Memory-30A3B model features approximately 3 billion activated parameters and attained an average score of 55.49, outperforming Opus5. The accompanying MemHub application connects with ten major AI tools, including Cursor and Codex, enabling real-time synchronization of personal memories across platforms. This technology is currently deployed in various business scenarios such as Tencent Stock, Tencent News, and iWiki.
Sources & editorial note
This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
- 腾讯混元Source published
Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.