Video understanding
LVBenchagentic
The same long-video test, run with the AI free to navigate the video itself — skipping around, replaying sections — rather than being shown it once. Read it as a separate test from the unaided run above. Higher is better.
Rankings
Higher is betterLVBench — frequently asked questions
- What is LVBench?
- The same long-video test, run with the AI free to navigate the video itself — skipping around, replaying sections — rather than being shown it once. Read it as a separate test from the unaided run above. Higher is better.
- Which AI model scores highest on LVBench?
- Gemini 3.8 Flash by Google holds the best LVBench (agentic) result among tracked models, at 87.8% (released Sep 2 2026). Higher scores are better on this benchmark.
- What are the top 1 models on LVBench?
- 1. Gemini 3.8 Flash (Google) — 87.8%.
- How many models have a published LVBench score?
- 1 tracked model has a published LVBench (agentic) score. Scores are the figures reported by each lab at that model's release, so this page is a record of results over time rather than a re-run leaderboard.