ShredBench - 8 Fragments: leaderboard

Document restoration on ShredBench (2026): news text in two languages, source code and tables are rendered, shredded into 8 irregular, physically deformed fragments and scattered, and the model restores the content from the fragment images. Normalized edit distance to the original from 0 to 1; lower is better.

Metric: Normalized edit distance (0-1) between the restored and the original document content, aggregated over all ShredBench categories (English and Chinese news, C++, Java and Python code, tables; 756 documents) when each page is shredded into 8 irregular Voronoi fragments rendered with 3D paper deformation and scattered; the model restores the content from the fragment images; temperature 0; lower is better. Source: arxiv.org. Saturation forecast: Around February 2027. 14 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3 Pro0.33#145
2Gemini 3 Flash0.34#175
3Qwen-VL-Plus0.59#5818
4Qwen-VL-Flash0.59
5GLM-4.6V0.67#710
6InternVL3.5-38B0.74#1021
7InternVL3.5-14B0.76#1504
8ShredBench Mistral3-Reas-14B (checkpoint unspecified)0.76
9GPT-5.10.77#240
10ShredBench Mistral3-Reas-8B (checkpoint unspecified)0.77
11InternVL3.5-8B0.78#1604
12GPT-5 Mini0.86#364
13DeepSeek-OCR0.86
14HunyuanOCR0.88

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

Interactive version: theaggregate.ai/benchmark?slug=shredbench-8-fragments · How It Works · Data refreshed daily, snapshot 2026-11-21.