Gemini 2.5
Google DeepMind — Gemini
rich detail 17 method fields published
Evaluation
Video understanding benchmark suite
Video understanding benchmarks (Table 6): string-match accuracy for multiple-choice VideoQA, LLM-based accuracy for open-ended VideoQA, R1@0.5 for moment retrieval, CIDEr for captioning; vs GPT 4.1 under comparable conditions
How it was run
Open a release to see the setup, scoring, and other details that the lab published. Only fields the lab actually disclosed are shown.