tau-Rec - Policy Compliance: leaderboard
Metric: Mean policy sub-score (0-1 scaled to %), programmatic compliance with the seven recommendation policies (recommend tool, watch history, availability, age restriction, sponsorship disclosure, transparency, single recommendation); the agent recommends one movie from a 153-title TMDB catalog to a GPT-5 mini user simulator through dialogue and catalog tools on 60 tasks (constraints revealed up front, only on asking, or never), 4 trials per task, agent temperature 0 where supported; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 Mini | 96.3 |
| 2 | GPT-5.4 (Medium) | 95.8 |
| 3 | DeepSeek V4 Flash (Max) | 94.2 |
| 4 | DeepSeek V4 Flash (High) | 93.7 |
| 5 | DeepSeek V4 Flash (Non-reasoning) | 92.1 |
| 6 | Gemini 2.5 Flash | 92 |
| 7 | GPT-5.4 (Non-reasoning) | 90.8 |
| 8 | Claude Sonnet 4.6 | 89.6 |
| 9 | Qwen 3 32B | 75.6 |
Interactive version: theaggregate.ai/benchmark?slug=tau-rec-policy-compliance · How It Works · Data refreshed daily, snapshot 2026-09-29.