Primary-research comparison, not our model execution
Memory writing: what to retain and update →
Hy-Memory-30B-A3B tackles a different step: writing conversation or work into memory that can help later.
Tencent Hunyuan / 腾讯混元 · Released 2026-10-15
Hy-Memory-30B-A3B
A specialist long-term-memory writer with about 30B total and 3B active parameters, organizing and updating memory inside external frameworks. Retrieval and final answers remain separate. Vendor comparisons on fixed 20% subsamples across four harnesses differ from AA text intelligence, Pi coding and Physics video continuation; no same-identity score currently qualifies it for those boards.
Original model explanation ↗No comparable leaderboard score is included yet. Explore the evaluations and examples below.
Evaluations: conditions and results
Vendor tests · fixed 20% subsamples · not our execution
Model card: the write path and four-harness evaluation ↗
Primary research · v1 · October 15, 2026
Original paper: training, attribution and evaluation scope ↗
Official repository · framework distinct from writer weights
Framework and evaluation utilities ↗
Tencent Hunyuan / 腾讯混元 model evolution
- Hy-Memory-30B-A3B · 研究与模型资料
Long-term memory-writing research learns what to retain, update and consolidate within memory frameworks. Its four-harness results use fixed 20% benchmark subsamples and do not establish a general-capability rank.
Original release ↗