List your company

Zero-shot video understanding on VideoZeroBench

27.6Level 3 ScoreGPT-5.4 xHigh
4.09610.19816.322.402May 30, 2026

Evaluation results

MethodLinks
GPT-5.4 xHighHarness=Codex, Reasoning effort=xHigh2026.0527.6
GPT-5.4 HighHarness=Codex, Reasoning effort=High2026.0527.2
GPT-5.4 MediumHarness=Codex, Reasoning effort=Medium2026.0524
GPT-5.4 LowHarness=Codex, Reasoning effort=Low2026.0521
GPT-5.4 High§Harness=Codex, Direct image read=disabled2026.0519
Gemini 3 Flash2026.0517.8
Gemini 3.1 Pro2026.0517.6
Kimi K-2.5Harness=CC2026.0515.6
OmniAgentBackbone=GPT-5.4 High2026.0512.2
Qwen 3.5 Omni Plus2026.0510.4
Mimo V2 Omni2026.057.4
MiniMax-M2.7‡Harness=CC2026.055.8
Qwen 3 Omni2026.055.2
Agent-OmniBackbone=GPT-5.4 High2026.055