MMCL-Bench: leaderboard
We introduce MMCL-Bench, a benchmark for multimodal context learning: learning task-local rules, procedures, and empirical patterns from visual or mixed-modality teaching context and applying them to new visual instances. Unlike text-only context learning or standard multimodal question answering, this setting requires models to recover and localize relevant evidence from images, screenshots, manu
Metric: Overall (self-reported). Source: benchmarklist.com. Saturation forecast: Around August 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 (Thinking) | 26.5 | #79 (GPT-5.4) |
| 2 | Qwen 3.6 Plus (Thinking) | 20.6 | #120 (Qwen 3.6 Plus) |
| 3 | Claude Opus 4.6 (Thinking) | 19.6 | #63 (Claude Opus 4.6) |
| 4 | Kimi K2.5 (Thinking) | 16.7 | #144 (Kimi K2.5) |
| 5 | Gemini 3.1 Pro (Preview) | 15.7 | #58 |
| 6 | GPT-5.4 | 14.7 | #79 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mmcl-bench · How It Works · Data refreshed daily, snapshot 2026-10-20.