MemLens (128K): leaderboard
MemLens tests the multimodal long-term memory of vision-language models with 789 questions whose answers depend on images and facts spread over many sessions. Overall accuracy in percent at a 128K-token interleaved image and text conversation history, with the evidence sessions hidden among distractor sessions (MemLens, 2026), judged by Qwen3-VL-235B; higher is better.
Metric: Overall accuracy (%) on all 789 questions (information extraction, multi-session, temporal and knowledge-update reasoning and answer refusal) at a 128K-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 February 2027. 25 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 51.99 | #99 |
| 2 | Kimi K2.5 | 51.99 | #266 |
| 3 | GPT-5.4 | 49.56 | #135 |
| 4 | Qwen 3 VL 235B A22B Instruct | 47.66 | #532 |
| 5 | Qwen 3.5 122B A10B | 45.5 | #346 |
| 6 | Qwen 3 VL 30B A3B Instruct | 44.23 | #895 |
| 7 | Qwen 3.5 27B | 40.68 | #375 |
| 8 | GLM-4.6V | 39.04 | #710 |
| 9 | Qwen 3 VL 8B Instruct | 34.73 | #1029 |
| 10 | Qwen 3 VL 235B A22B (Thinking) | 34.73 | #459 |
| 11 | Qwen 3.5 9B | 32.57 | #828 |
| 12 | Cosmos-Reason2-8B | 31.94 | #755 |
| 13 | Qwen 3 VL 30B A3B (Thinking) | 31.31 | #719 |
| 14 | Qwen 3 VL 8B (Thinking) | 30.29 | #910 |
| 15 | Qwen 3 VL 4B Instruct | 28.14 | #1332 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=memlens-128k · How It Works · Data refreshed daily, snapshot 2026-11-21.