MEGA-Bench Task - Number Puzzle Kakuro 5x5: leaderboard

Metric: Task Score (%). Source: huggingface.co. 44 models tracked.

Top models

#ModelScore
1Gemma 3 27B (IT)0
2GPT-4o0
3Gemma 3 12B (IT)0
4Gemma 3 4B (IT)0
5InternVL3-78B0
6InternVL3-8B0
7InternVL2-8B0
8Qwen 2 VL 72B0
9Qwen 2 VL 7B0
10Claude 3.5 Sonnet (20241022)0
11InternVL3-38B0
12Qwen 2 VL 2B0
13Aquila-VL-2B0
14GPT-4o Mini0
15Gemini 1.5 Pro (002)0

Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-number-puzzle-kakuro-5x5 · How It Works · Data refreshed daily, snapshot 2026-09-05.