EYT-Bench (PersonaMem-v2) - Persona Alignment: leaderboard

Metric: Persona alignment (style, value, culture, memory, security) (0-100): judge rubric of five sub-indicators scored 0, 0.5 or 1 per turn, summed to 0-5, turn weighted with the first two turns at 0.10 each and rescaled to 0-100; multi-turn dialogues of up to 10 turns with a persona-grounded Gemini 3.1 Pro user simulator and a Gemini 3.1 Pro judge (reasoning effort high, temperature 0); PersonaMem-v2 pool (100 dialogues; personas are free-text paragraphs distilled from long real user-assistant interactions). Source: arxiv.org. Saturation forecast: Estimated already saturated. 17 models tracked.

Top models

#ModelScore
1Gemini 3.1 Pro (Preview) (Thinking)95.7
2Qwen 3.5 27B95.3
3Claude Opus 4.794.3
4DeepSeek V4 Flash94.2
5Qwen 3.5 397B A17B93.3
6DeepSeek V4 Pro93.2
7Qwen 3.5 35B A3B92.9
8Seed 2.0 Lite92.7
9Claude Sonnet 4.691.3
10Seed 2.0 Pro88
11GPT-5.586.7
12Seed 2.0 Mini82.9

Interactive version: theaggregate.ai/benchmark?slug=eyt-bench-personamem-v2-persona-alignment · How It Works · Data refreshed daily, snapshot 2026-09-29.