Multimodal Reward Modeling on RewardBench Multimodal
99.6Safety ScoreR1-Reward
Evaluation results
| Method | Links | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| R1-RewardParams.=7B, Category=Open-Source Multimodal Reward Models2026.03 | 99.6 | 74.9 | 86.5 | 58.8 | 80 | — | — | — | — | — | — | — | — | — | — | |
| MSRL + voting@16Params.=8B, Backbone=InternVL3.5-8B, Voting Strategy=voting@162026.03 | 96.7 | 74.3 | 89.2 | 65.1 | 81.3 | — | — | — | — | — | — | — | — | — | — | |
| MSRLParams.=8B, Backbone=InternVL3.5-8B2026.03 | 96.4 | 72.8 | 88.6 | 64.2 | 80.5 | — | — | — | — | — | — | — | — | — | — | |
| Gemini-3-ProParams.=-, Category=MLLM-as-a-Judge2026.03 | 94.7 | 94.2 | 93 | 82.4 | 91.1 | — | — | — | — | — | — | — | — | — | — | |
| Gemini-1.5-ProParams.=-, Category=MLLM-as-a-Judge2026.03 | 94.5 | 66.3 | 87.2 | 63.9 | 63 | — | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL-8B Instruct + DeltaRubricModel=Qwen3-VL-8B Instruct, Rubric Strategy=DeltaRubric2026.05 | 93.3 | 69.5 | 84.9 | — | 73.2 | — | 65.9 | — | — | — | — | — | 76.9 | 68.7 | 52.6 | |
| MM-RLHF-RewardParams.=7B, Category=Open-Source Multimodal Reward Models2026.03 | 92.9 | 54.3 | 76.8 | 64.7 | 72.2 | — | — | — | — | — | — | — | — | — | — | |
| MM-RLHF-Reward#Param=7B2026.04 | 92.9 | — | — | — | 67.1 | 61.7 | 67.5 | 54.3 | 58.4 | 57.9 | — | — | — | — | — | |
| MM-RLHF-Reward-7BModel=MM-RLHF-Reward-7B2026.05 | 92.9 | 54.3 | 76.8 | — | 67.1 | — | 67.5 | — | — | — | — | — | 61.7 | 58.4 | 57.9 | |
| Qwen3-VL-8B Instruct + No rubricModel=Qwen3-VL-8B Instruct, Rubric Strategy=No rubric2026.05 | 91.5 | 56.7 | 77 | — | 68.7 | — | 62.7 | — | — | — | — | — | 75 | 64 | 51.5 | |
| Qwen3-VL-4B Instruct + DeltaRubricModel=Qwen3-VL-4B Instruct, Rubric Strategy=DeltaRubric2026.05 | 91.2 | 60 | 80.8 | — | 69.1 | — | 65.4 | — | — | — | — | — | 73.7 | 58 | 52 | |
| InternVL3.5-8BParams.=8B, Category=MLLM-as-a-Judge2026.03 | 90.6 | 66.5 | 83.4 | 62.6 | 75.8 | — | — | — | — | — | — | — | — | — | — | |
| IXC-2.5-Reward#Param=7B2026.04 | 89.9 | — | — | — | 66.6 | 60.7 | 64.2 | 56.8 | 63 | 50.5 | — | — | — | — | — | |
| Discriminative MRMParams.=8B, Backbone=InternVL3.5-8B2026.03 | 88.6 | 68.4 | 84.2 | 58.5 | 74.9 | — | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL-4B Instruct + No rubricModel=Qwen3-VL-4B Instruct, Rubric Strategy=No rubric2026.05 | 87.6 | 59.4 | 71.3 | — | 66.4 | — | 60.1 | — | — | — | — | — | 74.5 | 54.3 | 55 | |
| MSRL + voting@16Params.=4B, Backbone=InternVL3.5-4B, Voting Strategy=voting@162026.03 | 87.2 | 70.7 | 84.7 | 56.8 | 74.9 | — | — | — | — | — | — | — | — | — | — | |
| MSRLParams.=4B, Backbone=InternVL3.5-4B2026.03 | 85.4 | 69.8 | 84.3 | 56.2 | 73.9 | — | — | — | — | — | — | — | — | — | — | |
| Generative MRMParams.=8B, Backbone=InternVL3.5-8B2026.03 | 84.3 | 67.6 | 87.3 | 65.7 | 76.2 | — | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL-8B InstructModel=Qwen3-VL-8B Instruct, Rubric Strategy=Base2026.05 | 82.6 | 56.2 | 71.4 | — | 67.7 | — | 61.5 | — | — | — | — | — | 68.9 | 64.6 | 49.6 | |
| Qwen3-VL-8B-InstructParams.=8B, Category=MLLM-as-a-Judge2026.03 | 82.3 | 69.7 | 81.3 | 60.5 | 73.5 | — | — | — | — | — | — | — | — | — | — | |
| UnifiedReward-think (w/ CoT)Params.=7B, Category=Open-Source Multimodal Reward Models2026.03 | 81.5 | 63.4 | 78.5 | 66.4 | 72.5 | — | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL-4B InstructModel=Qwen3-VL-4B Instruct, Rubric Strategy=Base2026.05 | 80.4 | 52.7 | 70.8 | — | 65.3 | — | 56.3 | — | — | — | — | — | 66.1 | 46.7 | 54.4 | |
| Discriminative MRMParams.=4B, Backbone=InternVL3.5-4B2026.03 | 80.3 | 64.2 | 85.8 | 54.6 | 71.2 | — | — | — | — | — | — | — | — | — | — | |
| InternVL3.5-14BParams.=14B, Category=MLLM-as-a-Judge2026.03 | 79.9 | 73.5 | 72.4 | 67.8 | 73.4 | — | — | — | — | — | — | — | — | — | — | |
| LLaVA-Critic#Param=7B2026.04 | 78 | — | — | — | 54.1 | 50.1 | 53.2 | 51.3 | 49.2 | 49.3 | — | — | — | — | — | |
| Generative MRMParams.=4B, Backbone=InternVL3.5-4B2026.03 | 77.2 | 65.7 | 82.8 | 51.5 | 69.3 | — | — | — | — | — | — | — | — | — | — | |
| Claude 3.5 Sonnet#Param=-2026.04 | 76.8 | — | — | — | 71.5 | 62.6 | 67.8 | 73.9 | 68.6 | 65.1 | — | — | — | — | — | |
| GPT-4oParams.=-, Category=MLLM-as-a-Judge2026.03 | 74.8 | 72 | 87.2 | 65.4 | 74.9 | — | — | — | — | — | — | — | — | — | — | |
| GPT-4o#Param=-2026.04 | 74.8 | — | — | — | 70.8 | 62.6 | 69 | 72 | 67.6 | 62.1 | — | — | — | — | — | |
| Claude-3.7-SonnetParams.=-, Category=MLLM-as-a-Judge2026.03 | 72 | 78.1 | 86.8 | 66.7 | 75.9 | — | — | — | — | — | — | — | — | — | — | |
| Claude 3.7 Sonnet#Param=-2026.04 | 72 | — | — | — | 71.9 | 58.4 | 60.7 | 78.1 | 76.3 | 71.3 | — | — | — | — | — | |
| LLaVA-CriticParams.=7B, Category=Open-Source Multimodal Reward Models2026.03 | 70.7 | 56 | 69.4 | 59.7 | 64 | — | — | — | — | — | — | — | — | — | — | |
| UnifiedRewardParams.=7B, Category=Open-Source Multimodal Reward Models2026.03 | 64.7 | 58.8 | 68.5 | 56.4 | 62.1 | — | — | — | — | — | — | — | — | — | — | |
| Skywork-VL-Reward#Param=7B2026.04 | 62 | — | — | — | 65.7 | 64.5 | 62.4 | 54.3 | 70.8 | 44.7 | — | — | — | — | — | |
| InternVL3-8BModel=InternVL3-8B2026.05 | 59.3 | 60.5 | 82.3 | — | 63.6 | — | 61.6 | — | — | — | — | — | 59.6 | 65.1 | 56.6 | |
| VITA-1.5-7BModel=VITA-1.5-7B2026.05 | 58.1 | 52.5 | 50 | — | 53.6 | — | 54.3 | — | — | — | — | — | 55.6 | 51.9 | 52.8 | |
| UnifiedReward-Think#Param=7B2026.04 | 55.3 | — | — | — | 66.7 | 63.9 | 68.7 | 61.1 | 65.8 | 57.7 | — | — | — | — | — | |
| Llama-3.2-90BModel=Llama-3.2-90B2026.05 | 52 | 61.2 | 77.1 | — | 61.2 | — | 68.4 | — | — | — | — | — | 60 | 56.3 | 53.1 | |
| VL-MDR#Param=7B2026.04 | 51.2 | — | — | — | 69 | 69.8 | 68.7 | 67.8 | 78.2 | 45.2 | — | — | — | — | — | |
| GPT-5-miniParams.=-, Category=MLLM-as-a-Judge2026.03 | 50.4 | 76.8 | 51.2 | 70.6 | 62.3 | — | — | — | — | — | — | — | — | — | — | |
| UnifiedReward#Param=7B2026.04 | 50.4 | — | — | — | 61.6 | 63.1 | 58.4 | 55.9 | 64.4 | 44.7 | — | — | — | — | — | |
| Molmo-7BModel=Molmo-7B2026.05 | 34.8 | 54.6 | 60.3 | — | 52.9 | — | 59.4 | — | — | — | — | — | 56.8 | 50.7 | 53.4 | |
| Llama-3.2-11BModel=Llama-3.2-11B2026.05 | 20.9 | 55.5 | 55.8 | — | 51.2 | — | 65.8 | — | — | — | — | — | 57.8 | 50.6 | 51.7 | |
| SliME-8BModel=SliME-8B2026.05 | 19.1 | 47.5 | 53.8 | — | 42 | — | 52.2 | — | — | — | — | — | 42.3 | 43.5 | 35.3 | |
| BaseReward2026.04 | — | — | — | — | — | — | — | — | — | — | — | 72.8 | — | — | — | |
| Claude-Sonnet-4.52026.04 | — | — | — | — | — | — | — | — | — | — | 78.2 | — | — | — | — | |
| DT2IT-MRM2026.04 | — | — | — | — | — | — | — | — | — | — | — | 79.3 | — | — | — | |
| Gemini-2.5-Pro2026.04 | — | — | — | — | — | — | — | — | — | — | 82.4 | — | — | — | — | |
| GPT-52026.04 | — | — | — | — | — | — | — | — | — | — | 64.6 | — | — | — | — | |
| GPT-5.22026.04 | — | — | — | — | — | — | — | — | — | — | — | 75.3 | — | — | — | |
| InternVL3-78BSize=78B2026.04 | — | — | — | — | — | — | — | — | — | — | 75.7 | — | — | — | — | |
| InternVL3-8BSize=8B2026.04 | — | — | — | — | — | — | — | — | — | — | 66.9 | — | — | — | — | |
| IXC-2.5-RewardSize=7B2026.04 | — | — | — | — | — | — | — | — | — | — | 66.6 | — | — | — | — | |
| LLaVA-CriticSize=7B2026.04 | — | — | — | — | — | — | — | — | — | — | 62.2 | — | — | — | — | |
| MM-RLHF-RewardSize=7B2026.04 | — | — | — | — | — | — | — | — | — | — | 67.1 | — | — | — | — | |
| Molmo2-4BSize=4B2026.04 | — | — | — | — | — | — | — | — | — | — | 61.8 | — | — | — | — | |
| Molmo2-4B Multi-response RMSize=4B2026.04 | — | — | — | — | — | — | — | — | — | — | 73.2 | — | — | — | — | |
| Molmo2-8BSize=8B2026.04 | — | — | — | — | — | — | — | — | — | — | 66.8 | — | — | — | — | |
| Qwen2.5-VL-7BSize=7B2026.04 | — | — | — | — | — | — | — | — | — | — | 62.6 | — | — | — | — | |
| Qwen3-VL-32BSize=32B2026.04 | — | — | — | — | — | — | — | — | — | — | 79 | — | — | — | — | |
| Qwen3-VL-4BSize=4B2026.04 | — | — | — | — | — | — | — | — | — | — | 65.9 | — | — | — | — | |
| Qwen3-VL-4B Multi-response RMSize=4B2026.04 | — | — | — | — | — | — | — | — | — | — | 71.2 | — | — | — | — | |
| Qwen3-VL-8BSize=8B2026.04 | — | — | — | — | — | — | — | — | — | — | 71.6 | — | — | — | — | |
| R1-RewardSize=7B2026.04 | — | — | — | — | — | — | — | — | — | — | 82.2 | — | — | — | — | |
| Skywork-VL Reward2026.04 | — | — | — | — | — | — | — | — | — | — | — | 74.4 | — | — | — | |
| Skywork-VL-RewardSize=7B2026.04 | — | — | — | — | — | — | — | — | — | — | 74.2 | — | — | — | — |