List your company

Future Video Reasoning on FutureBench Average

85.4AccuracyFUTURE-L1
18.42435.81253.270.588Jun 4, 2026

Evaluation results

MethodLinks
FUTURE-L1Size=8B, Method=SFT+RL, Frames=322026.0685.4
Video-CoESize=7B, Method=SFT+RL, Frames=322026.0675
FUTURE-L1Size=8B, Method=SFT, Frames=322026.0673.2
Video-o3Size=7B, Method=SFT+RL, Frames=322026.0668.9
NEPSize=7B, Method=SFT+RL, Frames=322026.0667.3
Qwen3-VLSize=30B-A3B, Method=Zero-Shot, Frames=322026.0666.9
Text-Only SFTSize=8B, Method=SFT, Frames=322026.0665
VideoAuto-R1Size=8B, Method=SFT+RL, Frames=322026.0663.4
Video-R1Size=7B, Method=SFT+RL, Frames=322026.0663.3
SwimBirdSize=8B, Method=SFT, Frames=322026.0662.8
Qwen3-VL-InstructSize=8B, Method=Zero-Shot, Frames=322026.0661
GPT-4oSize=–, Method=Zero-Shot, Frames=322026.0659
Qwen2.5-VL-InstructSize=72B, Method=Zero-Shot, Frames=322026.0658.3
GPT-5Size=–, Method=Zero-Shot, Frames=322026.0657.9
InternVL3Size=8B, Method=Zero-Shot, Frames=322026.0656.7
Video-RFTSize=7B, Method=SFT+RL, Frames=322026.0654.6
Qwen2.5-VL-InstructSize=7B, Method=Zero-Shot, Frames=322026.0652.9
MiMo-VLSize=7B, Method=Zero-Shot, Frames=322026.0650.5
MonetSize=7B, Method=SFT+RL, Frames=322026.0647.9
LLaVA-NeXT-VideoSize=7B, Method=Zero-Shot, Frames=322026.0645.2
GLM-4.1VSize=9B, Method=Zero-Shot, Frames=322026.0644.4
LVRSize=7B, Method=SFT+RL, Frames=322026.0621