Hunyuan-T1-20250711: benchmark results

Provider: Tencent. Released 2025-07-11. Access: API.

Unified ELO 1784 ± 35, rank #179 of 2656 rated models, from 102 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
OpenCompass Language - Instruction Following - Chinese (CompassBench 2507)83.1Score (%)96.3
ReLE - Language and Instruction Following72.9Accuracy (%)93
OpenCompass Agent - Hard Multi-Turn (CompassBench 2507)50Score (%)91.7
OpenCompass Reasoning - Common - English (CompassBench 2507)75.6Score (%)91.7
SuperCLUE General (July 2025) - Hallucination Control83.39Score91.3
OpenCompass Code - Competition - Chinese (CompassBench 2507)59.4Score (%)88.9
OpenCompass Knowledge - Humanities - English (CompassBench 2507)87Score (%)88.9
OpenCompass Code - Competition - English (CompassBench 2507)60.9Score (%)87
SuperCLUE General (July 2025) - Precise Instruction Following42.73Score87
ReLE - Education - Primary School Subjects68Accuracy (%)85.4
OpenCompass LLM - Code (CompassBench 2507)43.8Score (%)85.2
OpenCompass LLM - Code - Chinese (CompassBench 2507)43.2Score (%)85.2

Interactive version: theaggregate.ai/model?slug=hunyuan-t1-20250711 · How It Works · Data refreshed daily, snapshot 2026-09-19.