Tencent Hunyuan Hy4 takes #1 Chinese model spot on Agent Arena at $0.57 per task
arena · x · 2026-11-06
Arena announces Tencent Hunyuan's Hy4 is now the top Chinese model on Agent Arena (#14 overall, only Chinese lab in the top 5), based on 12K+ real-world long-horizon agentic sessions.
- +5.4% net improvement, up 3.6 pts from the preview, with user-confirmed success at #11
- $0.57 median cost per task, firmly on the Pareto frontier — 62% cheaper than Kimi K3.1 Pro (Max) at $1.46/task with scores within 1 pt
- Currently proprietary; expected to tie Kimi K3.1 Pro as #1 open model once weights are released
Related event: Tencent Hunyuan Hy4 Tops Chinese Models on Agent Arena(2 posts)→
More from Models
- Claude hits conversation length limit after ~10 messages, user flags Anthropic — ___Patrice___ · 2026-11-06
- Researcher speculates on mystery model: LatentMoE, linear attention, 12T tokens — eliebakouch · 2026-11-06
- Yacine publicly asks insiders how many GPUs pretrained Inkling Small 2 — yacineMTB · 2026-11-06
- Autonomous AI R&D Runs 'Extremely Dumb Experiments', Researcher Reports — panickssery · 2026-11-06
- Claude Opus 5 pulls off multi-level roleplay: staying in character while knowing it's acting — repligate · 2026-11-06
- Inkling 2 Small hits 46.8% Pass@1 on APEX-Agents, up 13 points over predecessor — ziqiao_ma · 2026-11-06