FullHome: leaderboard
In real home deployments, household agents must often operate from a complete household scene and a situated household request, rather than from a clean task specification. Such requests require agents to identify task-relevant entities, recover intended task conditions, and resolve ordering constraints from the surrounding scene context. We formalize this capability as full-scene household reason
Metric: Goal SR (VirtualHome) (self-reported). Source: benchmarklist.com. Saturation forecast: Around April 2027. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 | 73.5 | #98 |
| 2 | DeepSeek V4 Flash | 63 | #137 |
| 3 | GPT-4.1 | 57 | #248 |
| 4 | GPT-4o | 49 | #343 |
| 5 | Gemini 2.5 Flash | 48.5 | #245 |
| 6 | Qwen 3.5 9B | 47.5 | #379 |
| 7 | MiMo-V2-Flash | 43.5 | #372 |
| 8 | Gemma 3 12B | 37 | #671 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=fullhome · How It Works · Data refreshed daily, snapshot 2026-10-20.