BALROG Crafter (VLM): leaderboard

Vision-language models on Crafter survival game using rendered screenshots, measuring visual interpretation for resource management and multi-step goals.

Metric: Progress (%). Source: balrogai.com. Saturation forecast: Estimated already saturated. 10 models tracked.

Top models

#ModelScoreOverall rank
1Gemini 2.5 Pro (Preview 03-25)37.27#123
2Claude 3.5 Sonnet (20241022)37.27#291
3Gemini 1.5 Pro (002)33.5#349
4GPT-4o (2024-05-13)26.81#325
5Gemini 1.5 Flash (002)20.7#583
6GPT-4o Mini (2024-07-18)19.91#558
7Llama 3.2 11B Instruct15.91#1126
8Llama 3.2 90B Instruct14.54#601
9Qwen 2 VL 7B Instruct5.45#826

No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.

Interactive version: theaggregate.ai/benchmark?slug=balrog-crafter-vlm · How It Works · Data refreshed daily, snapshot 2026-10-20.