MemLens (64K) - Knowledge Update: leaderboard
Metric: Accuracy (%) on the knowledge update (KU, 116 questions) at a 64K-token history, interleaved multimodal multi-session conversation history with the evidence sessions hidden among haystack sessions, judged correct or incorrect by a Qwen3-VL-235B-A22B-Instruct judge; higher is better. Source: arxiv.org. Saturation forecast: Around July 2027. 27 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Kimi K2.5 | 49.14 |
| 2 | Qwen 3.5 27B | 43.97 |
| 3 | GPT-5.4 | 43.1 |
| 4 | Qwen 3 VL 30B A3B Instruct | 39.66 |
| 5 | Qwen 3.5 9B | 37.93 |
| 6 | Qwen 3 VL 235B A22B Instruct | 36.21 |
| 7 | Qwen 3 VL 235B A22B (Thinking) | 35.34 |
| 8 | GLM-4.6V | 34.48 |
| 9 | Nemotron Nano 12B v2 VL | 34.48 |
| 10 | Qwen 3 VL 8B (Thinking) | 30.17 |
| 11 | Qwen 3 VL 8B Instruct | 29.31 |
| 12 | Qwen 3 VL 30B A3B (Thinking) | 29.31 |
| 13 | Qwen 3.5 4B | 28.45 |
| 14 | Gemma 3 27B (IT) | 27.59 |
| 15 | Phi-4 Multimodal Instruct | 24.14 |
Interactive version: theaggregate.ai/benchmark?slug=memlens-64k-knowledge-update · How It Works · Data refreshed daily, snapshot 2026-10-07.