List your company

Multi-way preference ranking on EDITREWARD-BENCH

66.2Preference Score (K=2)REWARDHARNESS (Gemini-2.0-Flash)
24.194435.099746.00556.9103Sep 30, 2025Nov 5, 2025Dec 12, 2025Jan 18, 2026Feb 24, 2026Apr 2, 2026May 9, 2026

Evaluation results

MethodLinks
REWARDHARNESS (Gemini-2.0-Flash)Sub-Agent=Gemini-2.0-Flash2026.0566.245.313.5—
Gemini-2.5-FlashMethod Category=Proprietary Models2025.0958.6139.8612.1638.02
Gemini-2.5-FlashModel Type=Proprietary Models2026.0558.639.912.2—
Claude-Haiku-4.5Model Type=Proprietary Models2026.0557.930.77.4—
REWARDHARNESS (Qwen)Sub-Agent=Qwen2026.0557.946.710.8—
GPT-5Method Category=Proprietary Models2025.0957.5338.5112.8437.81
GPT-5Model Type=Proprietary Models2026.0557.538.512.8—
EditReward (Qwen)Model Type=Open-Source Models, Backbone=Qwen2026.05573610.8—
EDITREWARDBackbone=Qwen2.5-VL-7B2025.0956.993610.8136.78
EditReward (MiMo)Model Type=Open-Source Models, Backbone=MiMo2026.0556.542.711.5—
EDITREWARDBackbone=MiMo-VL-7B-SFT2025.0956.4542.6711.4938.42
Qwen2.5-VL-7BModel Type=Open-Source Models2026.0552.724.73.4—
Qwen2.5-VL-7B-InstMethod Category=Open-Source VLMs2025.0952.6924.673.3829.75
Gemini-2.0-FlashMethod Category=Proprietary Models2025.0952.4333.3313.5133.47
Gemini-2.0-FlashModel Type=Proprietary Models2026.0552.433.313.5—
Qwen2.5-VL-3B-InstMethod Category=Open-Source VLMs2025.0951.0720.272.7126.86
Qwen2.5-VL-32B-InstMethod Category=Open-Source VLMs2025.0950.5425.274.0528.72
Qwen2.5-VL-32BModel Type=Open-Source Models2026.0550.525.34.1—
MiMo-VL-7BModel Type=Open-Source Models2026.0549.530.49.5—
MiMo-VL-7B-SFT-2508Method Category=Open-Source VLMs2025.0949.4630.419.4631.19
GPT-4oModel Type=Proprietary Models2026.0545.727.37.3—
GPT-4oMethod Category=Proprietary Models2025.0945.6927.337.3128.31
RandomMethod Category=Baseline2025.0925.8111.331.3513.84