List your company

Video Question Answering on Video-MME (Short, Medium, Long)

81.8VQA Accuracy (Overall)GPT-5
51.43259.31667.275.084Mar 31, 2026

Evaluation results

MethodLinks
GPT-5LLM=-2026.0381.8———
Gemini-1.5-ProLLM=-2026.037581.774.367.4
GPT-4oLLM=-2026.0371.98070.365.3
TSPOLLM=7B2026.0365.5——54.7
Qwen2.5-VLLLM=7B2026.0365.1———
LLaVA-Video + FlexMemLLM=7B2026.0364.776.363.354.4
ByteVideoLLMLLM=14B2026.0364.674.462.956.4
LLaVA-VideoLLM=7B2026.0363.276.16152.4
LongVULLM=7B2026.0360.6——59.5
mPLUG-Owl3LLM=7B2026.0359.37057.750.1
TimeMarkerLLM=8B2026.0357.37154.446.4
Video-XLLLM=7B2026.0355.56253.249.2
LongVALLM=7B2026.0352.661.150.446.2