WorldMemArena: leaderboard
Metric: QA Correct (%; share of answers a GPT-5.4-mini judge marks correct against the gold answer and evidence, over long multi-session memory trajectories). Source: arxiv.org. Saturation forecast: Around December 2026. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 Mini | 58.27 |
| 2 | Gemini 3 Flash | 51.69 |
| 3 | Qwen 3.5 Plus | 51.05 |
| 4 | Claude Haiku 4.5 | 36.71 |
Interactive version: theaggregate.ai/benchmark?slug=worldmemarena · How It Works · Data refreshed daily, snapshot 2026-09-26.