Tencent Hy4 Tops CWE-Bench-AA, Ranks #3 on Artificial Analysis Cyber Index
ArtificialAnlys · x · 2026-11-13
Artificial Analysis' new Cyber Index (combining CWE-Bench-AA, DeepsecBench-AA, and CyberGym-E2E-AA) ranks Tencent Hunyuan Hy4 third at 58%, behind Kimi K3.1 Pro (64%) and trusted-access GPT-6 Sol (69%), one point above DeepSeek V4.1 Flash 1017.
Key results:
- #1 on CWE-Bench-AA (76%), leading DeepSeek V4.1 Flash 1017 by 6 points on real-world vulnerability finding/fixing
- Strong cost efficiency: $0.92 per task, on the Capability vs. Cost Pareto frontier — 2 points above Grok 4.7 (xhigh) at 1/11th the cost
- Fewer output tokens: 64k per task, 43% less than DeepSeek V4.1 Flash 1017 and 37% less than Grok 4.7
API pricing: $0.834/$2.501 per 1M input/output tokens.
Related event: Tencent Hunyuan Hy4 Ranks Third in New Cyber Defense Index(2 posts)→
More from Models
- Opus 5.5 generates a full Blender 3D model in seconds, wows game developer — Scobleizer · 2026-11-13
- Claude's JavaScript art demo draws praise from threepointone: 'just remarkable' — threepointone · 2026-11-13
- Leaked UI shows unreleased Google models 'Hydrogen' and 'Argon' alongside Gemini tiers — Gauri_the_great · 2026-11-13
- DeepMind researcher: AI Overviews is the world's most-used AI model — cgarciae88 · 2026-11-13
- Claude Fable 5.2 Scores 41.3% on RLI, Nearly Double Rivals, as Sheets and Dashboards Roll Out — rubenhassid · 2026-11-13
- Gemini 4 Hydrogen hits AoE2 Bench's best quadrant: 44:32 win at $1.81 — sebkrier · 2026-11-13