MemLens (32K) - Multi-Session Reasoning: leaderboard

Accuracy in percent on the MemLens 143 questions that combine evidence from several sessions, at a 32K-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: Accuracy (%) on the multi-session reasoning (MSR, 143 questions) at a 32K-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: Not forecast. 26 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Kimi K2.544.06#266
2Gemini 3.1 Pro (Preview)32.17#99
3Qwen 3.5 122B A10B30.07#346
4Qwen 3.5 27B29.37#375
5Qwen 3 VL 235B A22B (Thinking)25.17#459
6Gemma 3 27B (IT)23.78#1324
7Cosmos-Reason2-8B23.78#755
8Gemma 3 12B (IT)22.38#1887
9Gemma 3 4B (IT)21.68#4139
10Qwen 3 VL 4B (Thinking)20.98#1334
11GLM-4.6V20.28#710
12Phi-4 Multimodal Instruct20.28#4154
13Qwen 3 VL 235B A22B Instruct18.88#532
14Qwen 3 VL 30B A3B Instruct18.88#895
15Qwen 3 VL 8B Instruct17.48#1029

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=memlens-32k-multi-session-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-21.