List your company

Long Video Understanding on CG-Bench long

38.6Accuracy (long-acc)InternVL3-8B
30.59232.67134.7536.829Apr 13, 2026

Evaluation results

MethodLinks
InternVL3-8BNum Frame=-, Token/Frame=-, Total Num of Token=>12k2026.0438.6—
POINTS1.5-8B (baseline)Num Frame=64, Token/Frame=324, Total Num of Token=≈ 20K2026.0436.753.9
POINTS-Long (standby)Num Frame=64, Token/Frame=32, Total Num of Token=2048 (10%)2026.0435.753.8
POINTS-Long (focus)Num Frame=64, Token/Frame=324+32, Total Num of Token=≈ 22K2026.0435.454.4
POINTS-Long (standby)Num Frame=64, Token/Frame=16, Total Num of Token=1024 (5%)2026.0434.652.9
POINTS-Long (standby)Num Frame=64, Token/Frame=8, Total Num of Token=512 (2.5%)2026.0433.652.7
LLaVA-OV-7BNum Frame=64, Token/Frame=-, Total Num of Token=>12k2026.0430.9—