Sycophancy (Lechmazur) — leaderboard
Tests whether models maintain consistent judgments when presented with opposite first-person perspectives of the same dispute. Measures sycophancy rate - how often a model agrees with both opposing narrators.
Metric: Sycophancy rate % (lower is better). Source: github.com. Status: saturated. 32 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 0.5 |
| 2 | Gemini 3.5 Flash | 0.5 |
| 3 | Grok 4.3 | 0.5 |
| 4 | Qwen 3.5 397B A17B | 2 |
| 5 | MiMo-V2.5-Pro | 2.5 |
| 6 | Claude Opus 4.6 (Non-reasoning) | 2.5 |
| 7 | Gemini 3.1 Flash Lite (Preview) | 3 |
| 8 | MiniMax-M3 | 3.5 |
| 9 | GPT-5.5 (High) | 3.5 |
| 10 | Qwen 3.6 Max Preview | 4 |
| 11 | GLM-5.1 | 4.5 |
| 12 | Kimi K2.6 | 4.5 |
| 13 | Gemma 4 31B (Reasoning) | 4.5 |
| 14 | Claude Opus 4.7 (High) | 4.5 |
| 15 | DeepSeek V4 Pro | 5 |
Interactive version: theaggregate.ai/benchmark?slug=sycophancy-lechmazur · How the rankings work · Data refreshed daily, snapshot 2026-07-22.