Tencent Hunyuan open-sources Hy-Memory-30B-A3B, an agent memory model topping Claude Opus 5 on memory benchmarks
AdinaYakup · x · 2026-10-16
Tencent Hunyuan released Hy-Memory-30B-A3B, a memory model for AI agents:
- 30B MoE with 3B active parameters, Apache 2.0 license, live on Hugging Face
- Extracts important memories from conversations and automatically updates outdated ones
- Benchmark chart shows an overall macro-average of 55.49, beating DeepSeek V4.1 Flash (53.64), size-matched open Qwen3.6-35B-A3B (50.37), and slightly ahead of closed Claude Opus 5 (54.62); it also leads on most Mem0, MemOS, and EverMEMOS splits, hitting 56.89 on EverMEMOS held-out
Related event: Tencent Hunyuan Open-Sources Hy Memory 2.0: 30B MoE Model for Agent Memory(5 posts)→
More from coding & agent
- Vibe-coded iPhone Duo sniping bot built with Codex in 17 minutes, 41 tests passing — xiaohu · 2026-10-16
- Long-term multiplayer agent memory tool tops BEAM memory benchmark — CurieuxExplorer · 2026-10-16
- NativeCAD ships generative CAD over MCP with 177 tools, 94% of plain-English part requests pass — burhop · 2026-10-16
- Flux.ai imports STEP files and edits existing CAD geometry via natural language — burhop · 2026-10-16
- 50 tasks at once energizes me, says dev, as peer warns AI context-switching causes burnout — zack_overflow · 2026-10-16
- Founder wipes 30k lines of tech debt in one shot with Opus — lucasmeijer · 2026-10-16