SMH-Bench (Interactive Agent): leaderboard
Metric: Success rate (%), instance-weighted over 1,100 tasks in 7 capability categories; Environment-Interactive Agent: a ReAct loop that starts from partial observability and queries and controls the HomeEnv simulator through tools under a call budget; final home state checked against the goal with untouched devices preserved; clarification and query answers judged by GPT-5; higher is better. Source: arxiv.org. Saturation forecast: Around December 2026. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 85.2 |
| 2 | Claude Sonnet 4.6 | 84.1 |
| 3 | GPT-5.4 | 79.6 |
| 4 | Qwen 3.5 397B A17B | 77.8 |
| 5 | Claude Haiku 4.5 | 76.2 |
| 6 | DeepSeek V3.2 (Thinking) | 75.5 |
| 7 | GPT-5.4 Mini | 75.3 |
| 8 | GLM-5 | 70.6 |
| 9 | Qwen 3.5 Plus | 70 |
| 10 | Qwen 3.5 9B | 64.3 |
| 11 | Qwen 3.5 4B | 56.5 |
Interactive version: theaggregate.ai/benchmark?slug=smh-bench-interactive-agent · How It Works · Data refreshed daily, snapshot 2026-09-29.