E-Bench (Internal)

Tencent internal (Hy team) office/enterprise-work benchmark, reported in the Hy4 preview model card appendix.

Category

General Agents

Max Score

100.0

Score Type

percent

Active

Yes

Model Scores

Scores for E-Bench (Internal)

Model Score Date Verified
Hy3
48.50
28.08.2026 Verified
Hy4 preview
89.10%
28.08.2026 Verified
DeepSeek V4 Pro
39.88%
28.08.2026 Verified
Qwen3.8-2.4T-A95B
57.01%
28.08.2026 Verified
GLM-5.3
71.34%
28.08.2026 Verified
Kimi K3
78.82%
28.08.2026 Verified
GPT-5.6 Sol
100.00%
28.08.2026 Verified
Claude Opus 5
91.28%
28.08.2026 Verified