Tencent Hy3 ranks 25th in Agent Arena after 8,000+ live sessions
arena · x · 2026-07-22
Tencent’s Hy3 ranks #25 in Agent Arena
Agent Arena shows Hy3 at #25 overall after 8K+ live agentic sessions.
- -2.2% net improvement overall
- -4.6% confirmed success
- -2.9% praise vs. complaint
- -7.1% steerability
- +2.6% bash recovery
- +0.9% tool hallucination
More from Models
- Kimi K3 trails leading U.S. frontier models in preliminary cyber tests — HZoete · 2026-07-24
- Commercial frontier models refused to analyze logs from a Hugging Face breach — garrytan · 2026-07-24
- Kimi K3 tops Frontend Code Arena and ties the best model in a designer blind test — yuwen_lu_ · 2026-07-24
- User says Opus 5 limits are so bad they’d pay 3x more to switch to Chinese models — MarcJSchmidt · 2026-07-24
- Google adds Computer Use support to Gemini 3.6 Flash and 3.5 Flash-Lite — patloeber · 2026-07-24
- Mathematician Claims LLMs Will Expose Hidden Gaps in Celebrated Academic Theorems — kylekabasares · 2026-07-24