DRInQ: leaderboard
Human conversation relies heavily on conversational implicature, in which speakers convey meanings that are suggested rather than explicitly stated. Although recent large language models exhibit strong conversational fluency, they remain unreliable when interpretation depends on reasoning that integrates social and contextual cues, a process rarely articulated in text. We introduce DRinQ, a benchm
Metric: Vanilla (self-reported). Source: benchmarklist.com. Saturation forecast: Around February 2027. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Median Human | 88 | |
| 2 | O3 | 67 | #124 |
| 3 | Claude Sonnet 4.5 | 65 | #142 |
| 4 | Claude Haiku 4.5 | 64 | #280 |
| 5 | GPT-4o | 62 | #343 |
| 6 | GPT-5 | 62 | #98 |
| 7 | DeepSeek R1 | 62 | #251 |
| 8 | GPT-4.1 | 61 | #248 |
| 9 | DeepSeek V3 | 60 | #321 |
| 10 | Llama 3.3 70B | 58 | #579 |
| 11 | GPT-5 Mini | 56 | #186 |
| 12 | Qwen 2.5 72B | 56 | #499 |
| 13 | GPT-5 Nano | 45 | #425 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=drinq · How It Works · Data refreshed daily, snapshot 2026-10-20.