Baba Is Harbor - Stage 0: leaderboard

Tests agents playing Baba Is You across eight introductory levels using the same harness for every model. Reports the percentage of levels solved in stage 0; later stages using different model-specific harnesses are excluded from this series.

Metric: pass@1 (%). Source: quesma.com. Saturation forecast: Around February 2027. 29 models tracked.

Top models

#ModelScoreOverall rank
1GPT-5.5100#33
2GPT-5.6 Sol100#22
3Claude Opus 5100#16
4Gemini 3.8 Flash100#30
5GPT-6 Astra100#6
6Claude Fable 5100#20
7Gemini 3.7 Flash100#24
8Claude Fable 5.1100#10
9Claude Opus 4.896#39
10Kimi K396#35
11GLM-5.296#80
12GLM-5.396#44
13Grok 4.696#29
14Gemini 3.1 Pro (Preview)92#64
15GPT-5.6 Terra88#40

No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #9 Gemini 4 Argon.

Interactive version: theaggregate.ai/benchmark?slug=baba-is-harbor-stage-0 · How It Works · Data refreshed daily, snapshot 2026-11-05.