List your company
ResearchFormerly WizWand

Image Reasoning on WeMath

71.3AccuracyGemini-2.5-Pro
8.58824.86941.1557.431Apr 29, 2026May 1, 2026May 4, 2026May 6, 2026May 9, 2026May 11, 2026May 14, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Gemini-2.5-ProModel Type=Closed-source2026.0571.3
Claude-4-SonnetModel Type=Closed-source2026.0563
OPD V→TTraining Strategy=Static On-policy Policy Distillation (Image to Text)2026.0460.95
MOPDTraining Strategy=MOPD2026.0459.9
CoPDTraining Strategy=CoPD2026.0459.9
CoPDTraining Strategy=Co-Evolving Policy Distillation2026.0459.81
Image-ExpertTraining Strategy=Image-Specific Expert2026.0459.14
Image-ExpertTraining Strategy=Image-Expert2026.0459.14
Mixed RLVRTraining Strategy=Mixed RLVR2026.0459.05
OPD T→VTraining Strategy=Static On-policy Policy Distillation (Text to Image)2026.0458.86
Mixed RLVRTraining Strategy=Joint optimization on combined budget2026.0457.43
Text-ExpertTraining Strategy=Text-Specific Expert2026.0455.24
Text-ExpertTraining Strategy=Text-Expert2026.0455.24
Video-ExpertTraining Strategy=Video-Expert2026.0453.43
BaseTraining Strategy=Base Model2026.0452.38
BaseTraining Strategy=Base2026.0452.38
GPT-4oModel Type=Closed-source2026.0550.6
Gemini-2.0-FlashModel Type=Closed-source2026.0547.4
ATLASLA-GRPOTraining=LA-GRPO2026.0545
VTS-VModel Type=Agentic Visual2026.0542.8
LVRModel Type=Latent Visual2026.0541.2
ATLASGRPOTraining=GRPO2026.0540.3
MCOTModel Type=Latent Visual2026.0539.6
BagelModel Type=Unified2026.0539.4
CoVTModel Type=Latent Visual2026.0538.1
MonetModel Type=Latent Visual2026.0536.9
Qwen2.5-VLModel Type=Standard2026.0536.2
V-ThinkerModel Type=Agentic Visual2026.0532.5
Gemma-3-27BModel Type=Standard2026.0531.7
ATLASSFTTraining=SFT2026.0528.9
Visual CoTModel Type=Agentic Visual2026.0528.6
AnoleModel Type=Unified2026.0524.7
LLaVA-OneVision-7BModel Type=Standard2026.0523.1
MiniGPT-v2Model Type=Standard2026.0511