MMCL-Bench - Procedural Task Execution: leaderboard
Metric: Strict task pass rate (%) on the 32 procedural tasks (manuals, specifications and workflows; a task passes only when every rubric item is satisfied), on the 102 MMCL-Bench tasks (49 rule-system, 32 procedural, 21 empirical), in which a model learns a task-local rule, procedure or pattern from visual or mixed-modality teaching context and applies it to a new visual instance, scored against expert rubrics by a rubric-only LLM judge, models at their recommended or default decoding settings; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 (Thinking) | 25 |
| 2 | Kimi K2.5 (Thinking) | 21.9 |
| 3 | Qwen 3.6 Plus (Thinking) | 18.8 |
| 4 | GPT-5.4 (Thinking) | 18.8 |
| 5 | Gemini 3.1 Pro (Preview) | 12.5 |
| 6 | GPT-5.4 | 9.4 |
Interactive version: theaggregate.ai/benchmark?slug=mmcl-bench-procedural-task-execution · How It Works · Data refreshed daily, snapshot 2026-10-07.