WDCD R3 Pressure Integrity: leaderboard
WDCD Round 3 — maintaining answer integrity under conversational pressure; round score.
Metric: Score (%). Source: huggingface.co. Saturation forecast: Not forecast. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.7 | 150 | #49 |
| 2 | DeepSeek V4 Pro | 138 | #102 |
| 3 | Gemini 2.5 Pro | 125 | #152 |
| 4 | Claude Sonnet 4.6 | 125 | #90 |
| 5 | GPT-6 Astra | 125 | #6 |
| 6 | Gemini 3.1 Pro (Preview) | 113 | #58 |
| 7 | Grok 4 | 113 | #181 |
| 8 | GPT-6 Sol | 113 | #14 |
| 9 | GPT-6.1 Sol | 100 | #7 |
| 10 | GLM-4.6 | 88 | #252 |
| 11 | GPT-6 Luna | 88 | #56 |
| 12 | GPT-5.5 | 50 | #31 |
| 13 | O3 | 50 | #124 |
| 14 | Qwen 3 Max | 0 | #208 |
No result here: #3 Claude Opus 5.5, #8 Gemini 4 Argon, #9 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=wdcd-r3-pressure-integrity · How It Works · Data refreshed daily, snapshot 2026-10-20.