tau-Rec - Constraint Score: leaderboard
Metric: Mean constraint sub-score (0-1 scaled to %), the partial-credit share of typed constraints the recommendation satisfies; 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: Around January 2027. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V4 Flash (Max) | 59.6 |
| 2 | DeepSeek V4 Flash (High) | 58.6 |
| 3 | DeepSeek V4 Flash (Non-reasoning) | 58.3 |
| 4 | Claude Sonnet 4.6 | 57.9 |
| 5 | GPT-5.4 (Medium) | 56.5 |
| 6 | GPT-5.4 (Non-reasoning) | 49.6 |
| 7 | GPT-5 Mini | 42.9 |
| 8 | Qwen 3 32B | 36.1 |
| 9 | Gemini 2.5 Flash | 31.1 |
Interactive version: theaggregate.ai/benchmark?slug=tau-rec-constraint-score · How It Works · Data refreshed daily, snapshot 2026-09-29.