List your company
ResearchFormerly WizWand

Visual Question Answering on RWQA

80.8AccuracyGPT-5
-3.23218.58440.462.216Jun 13, 2025Aug 14, 2025Oct 16, 2025Dec 18, 2025Feb 19, 2026Apr 23, 2026Jun 25, 2026
Updated 4mo ago

Evaluation results

MethodLinks
GPT-5Access=API call only2026.0380.8
VISEModel Scale=32B2026.0679.51
VisionZero-ChartModel Scale=32B2026.0679.34
VisionZero-CLEVRModel Scale=32B2026.0679.29
iReasonerModel Scale=32B2026.0679.05
EvoLMMModel Scale=32B2026.0678.98
VisionZero-RWModel Scale=32B2026.0678.31
VisPlayModel Scale=32B2026.0678.24
BaseModel Scale=32B2026.0678.16
Molmo2-8BAccess=Fully open2026.0377.6
Gemini 2.5 ProAccess=API call only2026.0377.4
MolmoPoint-8BAccess=MolmoPoint2026.0377.4
GPT-5 miniAccess=API call only2026.0377
Eagle2.5-8BAccess=Open weights only2026.0376.7
Molmo2-4BAccess=Fully open2026.0375.4
PLM-8BAccess=Fully open2026.0375
Qwen3-VL-8BTraining=Staged2026.0574.51
Gemini 2.5 FlashAccess=API call only2026.0374.5
MolmoPoint-8B-O-7BAccess=Fully open2026.0373.6
Keye-VL-1.5-8BAccess=Open weights only2026.0373.5
PLM-3BAccess=Fully open2026.0372.4
MiniCPM-V-4.5-8BAccess=Open weights only2026.0372.1
Qwen3-VL-8BAccess=Open weights only2026.0371.5
OneThinker-8BTraining=Base2026.0571.5
Qwen3-VL-4BAccess=Open weights only2026.0370.9
Qwen3-VL-8BTraining=Base2026.0570.85
GLM-4.1V-9BAccess=Open weights only2026.0370.7
Qwen2.5-VL-7BTraining=Staged2026.0570.46
VISEModel Scale=4B2026.0670.46
VISEModel Scale=8B2026.0670.03
VisionZero-ChartModel Scale=8B2026.0669.95
VisionZero-CLEVRModel Scale=8B2026.0669.88
iReasonerModel Scale=8B2026.0669.81
GThinker-7BTraining=Base2026.0569.8
VisionZero-RWModel Scale=4B2026.0669.74
EvoLMMModel Scale=8B2026.0669.74
WeThink-7BTraining=Base2026.0569.54
OpenVLThinker-7BTraining=Base2026.0569.41
VisPlayModel Scale=8B2026.0669.41
VisionZero-RWModel Scale=8B2026.0669.33
BaseModel Scale=8B2026.0669.28
iReasonerModel Scale=4B2026.0668.92
VisPlayModel Scale=4B2026.0668.67
EvoLMMModel Scale=4B2026.0668.65
VisionZero-CLEVRModel Scale=4B2026.0668.43
VisionZero-ChartModel Scale=4B2026.0667.91
BaseModel Scale=4B2026.0667.82
InternVL3.5-8BAccess=Open weights only2026.0367.5
Qwen2.5-VL-7BTraining=Base2026.0567.45
InternVL3.5-4BAccess=Open weights only2026.0366.3
VISEModel Scale=2B2026.0664.58
MMR1-7BTraining=Base2026.0564.31
VisionZero-RWModel Scale=2B2026.0664.31
iReasonerModel Scale=2B2026.0663.94
EvoLMMModel Scale=2B2026.0663.78
VisionZero-CLEVRModel Scale=2B2026.0663.75
VisionZero-ChartModel Scale=2B2026.0663.64
VisPlayModel Scale=2B2026.0663.62
InstructVLA-Generalistnumber of parameters=1.5B, robot state=true2025.0763.5
BaseModel Scale=2B2026.0663.41
R1-OneVision-RL-7BTraining=Base2026.0563.4
Eagle2number of parameters=1.5B2025.0763.1
InstructVLA-Generalistnumber of parameters=1.5B2025.0763.1
Qwen2-VLnumber of parameters=1.5B2025.0762.9
Claude Sonnet 4.5Access=API call only2026.0361.1
Bunnynumber of parameters=8B2025.0760.4
VGR-7BDownSample=2×2 | 4×4, #Vtoken=864, LLM=Vicuna-7B2025.0659.8
VGR-7BDownSample=2×2 | 2×2, #Vtoken=3024, LLM=Vicuna-7B2025.0659.5
LLaVA-NeXT-7BDownSample=No, #Vtoken=2880, LLM=Vicuna-7B2025.0657.8
ChatVLAnumber of parameters=1.5B2025.0757
LLaVA-NeXT-7B†DownSample=2×2 | 4×4, #Vtoken=864, LLM=Vicuna-7B2025.0656.8
Magmanumber of parameters=8B2025.0756.5
PaliGemmanumber of parameters=2B2025.0755.2
LLaVA-v1.5-7BDownSample=No, #Vtoken=576, LLM=Vicuna-7B2025.0654.8
DeepSeek-VL-7BDownSample=Conv2D, #Vtoken=576, LLM=DeepSeek-7B2025.0654.2
Qwen-VL-Chat-7BDownSample=Cross-Attn, #Vtoken=1024, LLM=Qwen-7B2025.0649.3
OpenVLAnumber of parameters=7B, fine-tuned=true2025.0747.2
ECoTnumber of parameters=7B2025.0729.8
OpenVLAnumber of parameters=7B2025.070