List your company

Video-demo In-context Learning on Demo-ICL-Bench

80.4Accuracy (Demo)Human
20.0835.7451.467.06Feb 9, 2026

Evaluation results

MethodLinks
Human2026.0280.4——
Gemini-2.5-Pro2026.0236.2——
Demo-ICLSize=7B, Frame=322026.023227.64.4
GPT-4o2026.0231.4——
LLaVA-VideoSize=7B, Frame=322026.0230.2291.2
Demo-ICL (SFT)Size=7B, Frame=322026.0229.426.23.2
VideoChat-R1Size=7B, Frame=322026.0228.226.81.4
Qwen2.5-VLSize=7B, Frame=322026.022826.21.8
Video-R1Size=7B, Frame=322026.0227.426.60.8
InternVL-3Size=8B, Frame=322026.022726.40.6
Qwen2.5-VLSize=72B, Frame=322026.0225.625.20.4
Ola-Video (Base)Size=7B, Frame=322026.022526-1
OlaSize=7B, Frame=322026.0224.626.4-1.8
Qwen2-VLSize=7B, Frame=322026.0222.424-1.6