List your company
ResearchFormerly WizWand

Future Video Reasoning on FutureBench 1-Hop

83.2AccuracyFUTURE-L1
20.07236.46152.8569.239Jun 4, 2026
Updated 5mo ago

Evaluation results

MethodLinks
FUTURE-L1Size=8B, Method=SFT+RL, Frames=322026.0683.2
Video-CoESize=7B, Method=SFT+RL, Frames=322026.0680.9
FUTURE-L1Size=8B, Method=SFT, Frames=322026.0670.5
Video-o3Size=7B, Method=SFT+RL, Frames=322026.0668.2
Video-R1Size=7B, Method=SFT+RL, Frames=322026.0667.6
Text-Only SFTSize=8B, Method=SFT, Frames=322026.0667.6
NEPSize=7B, Method=SFT+RL, Frames=322026.0666.2
Qwen3-VLSize=30B-A3B, Method=Zero-Shot, Frames=322026.0665.3
Qwen3-VL-InstructSize=8B, Method=Zero-Shot, Frames=322026.0664.2
VideoAuto-R1Size=8B, Method=SFT+RL, Frames=322026.0663.6
Video-RFTSize=7B, Method=SFT+RL, Frames=322026.0662.4
GPT-4oSize=–, Method=Zero-Shot, Frames=322026.0661.9
GPT-5Size=–, Method=Zero-Shot, Frames=322026.0659.6
MiMo-VLSize=7B, Method=Zero-Shot, Frames=322026.0659
SwimBirdSize=8B, Method=SFT, Frames=322026.0659
Qwen2.5-VL-InstructSize=7B, Method=Zero-Shot, Frames=322026.0657.2
Qwen2.5-VL-InstructSize=72B, Method=Zero-Shot, Frames=322026.0655.5
InternVL3Size=8B, Method=Zero-Shot, Frames=322026.0654.3
LLaVA-NeXT-VideoSize=7B, Method=Zero-Shot, Frames=322026.0648.8
MonetSize=7B, Method=SFT+RL, Frames=322026.0646.8
GLM-4.1VSize=9B, Method=Zero-Shot, Frames=322026.0629.9
LVRSize=7B, Method=SFT+RL, Frames=322026.0622.5