VideoGameBench Lite: leaderboard
Metric: Mean game progress (%; VG-Agent; paused during inference; 3 test games). Source: vgbench.com. Saturation forecast: Around 2030. 5 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 1.6 |
| 2 | GPT-4o (2024-08-06) | 1.6 |
| 3 | Claude 3.7 Sonnet (20250219) | 1.6 |
| 4 | Gemini 2.0 Flash | 0 |
| 5 | Llama 4 Maverick Instruct FP8 | 0 |
Interactive version: theaggregate.ai/benchmark?slug=videogamebench-lite · How It Works · Data refreshed daily, snapshot 2026-10-09.