List your company
ResearchFormerly WizWand

Open-ended Writing on ArenaHard

50AccuracyOfficial Instruct Model
-1.854411.607825.0738.5322May 31, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Official Instruct ModelEvaluation Category=Official Instruct Model2026.0550
Opus-4.7 (xHigh)Evaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.0533.84
GLM-4.7 & ANDESEvaluation Category=Proposed Method Comparison, Base Model=Qwen3-1.7B2026.0512.94
GLM-4.7 (Scaffold-only)Evaluation Category=Proposed Method Comparison, Base Model=Qwen3-1.7B2026.055.7
Opus-4.6 (1M)Evaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.053.21
Gemini-3.1-ProEvaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.052.27
MiniMax-M2.5Evaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.051.76
GPT-5.2Evaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.051.27
Opus-4.6Evaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.051.15
Base Model (Qwen3-1.7B)Evaluation Category=Zero-Shot2026.050.91
MiniMax-M2.1Evaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.050.91
Qwen3-MaxEvaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.050.91
Kimi-K2-ThinkingEvaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.050.91
GPT-5.4 (High)Evaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.050.91
GLM-4.7 (OpenCode)Evaluation Category=Proposed Method Comparison, Base Model=Qwen3-1.7B2026.050.91
Sonnet-4.5Evaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.050.21
GPT-5.1-Codex-MaxEvaluation Category=Agent-Post-Trained, Base Model=Qwen3-1.7B2026.050.14