hy4: benchmark results

Provider: Tencent. Released 2026-11-04. Access: API.

Aggregate Rating 1944 ± 48, rank #13 of 1635 rated models, from 16 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AI Chess Leaderboard (Continuation)1802Elo99.6
AI Chess Leaderboard (Reasoning)1810Elo98.9
FoodTruckBench67597.64Source-selected End Net Worth ($)98.3
Vals AI Tax Agent Bench34.4Accuracy (%)82.5
Vals AI Vibe Code Bench81.01Accuracy (%)77.2
Vals AI Terminal-Bench 4.039.39Accuracy (%)76.3
Vals AI Finance Agent v2.164.71Dealbreaker-gated Partial Credit (%)71.4
Vals AI Finance Index58.05Finance Index (%)68.6
Vals AI Legal Research Bench41.35Accuracy (%)68.5
Vals AI (Vals Index)59.35Accuracy (%)62.5
Vals AI Legal Index70.55Legal Index (%)60.6
Vals AI Harvey Legal Agent Bench v1.112.5Task Resolution (%)54.5

Interactive version: theaggregate.ai/model?slug=hy4 · How It Works · Data refreshed daily, snapshot 2026-11-05.