PEC-Home (ICL) - Multi-User Level 1: leaderboard

Metric: Execution accuracy (%; 1,068 smart-home commands at ellipsis level 1 of 4 (complete commands), functionally correct device operations; one in-context example per ellipsis level, no dialogue history). Source: arxiv.org. Saturation forecast: Estimated already saturated. 10 models tracked.

Top models

#ModelScore
1Gemma 2 27B (IT)93.63
2Gemma 2 9B (IT)93.26
3GPT-4o92.98
4DeepSeek V392.79
5Qwen 2.5 32B Instruct92.13
6Mistral 7B Instruct (v0.3)90.36
7Qwen 2.5 72B Instruct89.42
8Qwen 2.5 7B Instruct88.58
9Qwen 2.5 14B Instruct85.86
10Llama 3.1 8B Instruct79.21

Interactive version: theaggregate.ai/benchmark?slug=pec-home-icl-multi-user-level-1 · How It Works · Data refreshed daily, snapshot 2026-09-26.