IKEA-Bench - Progress Tracking: leaderboard
Metric: Accuracy (%) of progress tracking (locate the current step shown in video among all manual step diagrams of the product, four-way multiple choice; 334 questions; chance 25) on IKEA-Bench, wordless IKEA manual diagrams against real assembly video from 29 products, Visual setting (diagram images only), zero-shot greedy decoding, unparsed answers counted wrong; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 19 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Flash | 65.6 |
| 2 | Gemini 3.1 Pro (Preview) | 65 |
| 3 | Qwen 3.5 27B | 59.3 |
| 4 | Qwen 3.5 9B | 46.7 |
| 5 | GLM-4.1V-9B (Thinking) | 43.7 |
| 6 | Gemma 3 27B | 37.1 |
| 7 | Gemma 3 12B | 35.9 |
| 8 | Qwen 2.5 VL 7B | 35 |
| 9 | Qwen 3.5 2B | 32.9 |
| 10 | Gemma 3 4B | 27.8 |
Interactive version: theaggregate.ai/benchmark?slug=ikea-bench-progress-tracking · How It Works · Data refreshed daily, snapshot 2026-10-07.