Hugging Face's logo

Datasets:
MTSAIR
/
MWS-Vision-Bench

Tasks:
Visual Question Answering
Document Question Answering
Modalities:
Image
Text
Document
Formats:
parquet
Languages:
Russian
Size:
1K - 10K
Tags:
benchmark
multimodal
ocr
kie
grounding
vlm
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
3
sky-pirate-2020 commited on
Commit
04a1127
·
verified ·
1 Parent(s): 8f9e77a

Add October 2026 model results to leaderboard

Browse files

Adds the new rows from the canonical GitHub leaderboard (mts-ai/MWS-Vision-Bench@461b9de). Existing rows are unchanged.

Files changed (1) hide show
  1. README.md +8 -0
README.md CHANGED
@@ -237,21 +237,29 @@ Top models evaluated on this validation dataset:
237
 
238
  | Model | Overall | img→text | img→markdown | Grounding | KIE (JSON) | VQA | Anti-fraud |
239
  |-------|---------|----------|--------------|-----------|------------|-----|------------|
 
 
240
  | Claude Fable 5 | 0.799 | 0.828 | 0.737 | 0.635 | 0.843 | 0.950 | 0.521 |
241
  | GPT-5.6 Sol | 0.790 | 0.818 | 0.716 | 0.621 | 0.867 | 0.926 | 0.467 |
242
  | GPT-5.5 | 0.778 | 0.855 | 0.686 | 0.563 | 0.873 | 0.913 | 0.477 |
243
  | Kimi K3 | 0.777 | 0.783 | 0.716 | 0.588 | 0.856 | 0.941 | 0.400 |
 
244
  | GPT-5.6 Terra | 0.749 | 0.808 | 0.693 | 0.571 | 0.799 | 0.874 | 0.404 |
245
  | Kimi K2.6 | 0.743 | 0.868 | 0.761 | 0.301 | 0.858 | 0.929 | 0.279 |
246
  | Cotype Pro 3 | 0.742 | 0.868 | 0.754 | 0.423 | 0.873 | 0.791 | 0.233 |
247
  | Claude Sonnet 5 | 0.731 | 0.842 | 0.729 | 0.370 | 0.788 | 0.924 | 0.391 |
248
  | GPT-5.6 Luna | 0.727 | 0.776 | 0.673 | 0.547 | 0.772 | 0.864 | 0.302 |
 
249
  | Gemini 3.6 Flash | 0.723 | 0.833 | 0.736 | 0.178 | 0.903 | 0.967 | 0.115 |
 
 
250
  | Qwen 3.7 Plus | 0.708 | 0.843 | 0.766 | 0.154 | 0.844 | 0.935 | 0.250 |
251
  | Claude-4.6-Opus | 0.704 | 0.841 | 0.748 | 0.168 | 0.852 | 0.908 | 0.385 |
 
252
  | Qwen3.6-35B-A3B | 0.694 | 0.859 | 0.784 | 0.099 | 0.822 | 0.904 | 0.232 |
253
  | Qwen3.6-27B | 0.691 | 0.863 | 0.768 | 0.071 | 0.833 | 0.919 | 0.195 |
254
  | Gemini-2.5-pro | 0.690 | 0.840 | 0.717 | 0.070 | 0.888 | 0.935 | 0.088 |
 
255
  | Kimi K2 Instruct | 0.686 | 0.874 | 0.726 | 0.148 | 0.805 | 0.876 | 0.331 |
256
  | Cotype Light 3 | 0.685 | 0.855 | 0.770 | 0.125 | 0.873 | 0.799 | 0.318 |
257
  | GPT-5.4 | 0.683 | 0.764 | 0.627 | 0.351 | 0.826 | 0.846 | 0.432 |
 
237
 
238
  | Model | Overall | img→text | img→markdown | Grounding | KIE (JSON) | VQA | Anti-fraud |
239
  |-------|---------|----------|--------------|-----------|------------|-----|------------|
240
+ | GPT-6.1 Sol | 0.822 | 0.847 | 0.743 | 0.711 | 0.859 | 0.952 | 0.487 |
241
+ | Claude Opus 5.5 | 0.819 | 0.824 | 0.749 | 0.659 | 0.896 | 0.967 | 0.611 |
242
  | Claude Fable 5 | 0.799 | 0.828 | 0.737 | 0.635 | 0.843 | 0.950 | 0.521 |
243
  | GPT-5.6 Sol | 0.790 | 0.818 | 0.716 | 0.621 | 0.867 | 0.926 | 0.467 |
244
  | GPT-5.5 | 0.778 | 0.855 | 0.686 | 0.563 | 0.873 | 0.913 | 0.477 |
245
  | Kimi K3 | 0.777 | 0.783 | 0.716 | 0.588 | 0.856 | 0.941 | 0.400 |
246
+ | GPT-6 Luna | 0.770 | 0.817 | 0.734 | 0.581 | 0.820 | 0.897 | 0.400 |
247
  | GPT-5.6 Terra | 0.749 | 0.808 | 0.693 | 0.571 | 0.799 | 0.874 | 0.404 |
248
  | Kimi K2.6 | 0.743 | 0.868 | 0.761 | 0.301 | 0.858 | 0.929 | 0.279 |
249
  | Cotype Pro 3 | 0.742 | 0.868 | 0.754 | 0.423 | 0.873 | 0.791 | 0.233 |
250
  | Claude Sonnet 5 | 0.731 | 0.842 | 0.729 | 0.370 | 0.788 | 0.924 | 0.391 |
251
  | GPT-5.6 Luna | 0.727 | 0.776 | 0.673 | 0.547 | 0.772 | 0.864 | 0.302 |
252
+ | Qwen3.8 Flash | 0.726 | 0.842 | 0.739 | 0.301 | 0.833 | 0.913 | 0.353 |
253
  | Gemini 3.6 Flash | 0.723 | 0.833 | 0.736 | 0.178 | 0.903 | 0.967 | 0.115 |
254
+ | GLM-5.3 Flash | 0.709 | 0.651 | 0.606 | 0.603 | 0.811 | 0.874 | 0.420 |
255
+ | Gemini 3.8 Flash | 0.709 | 0.788 | 0.724 | 0.159 | 0.896 | 0.978 | 0.175 |
256
  | Qwen 3.7 Plus | 0.708 | 0.843 | 0.766 | 0.154 | 0.844 | 0.935 | 0.250 |
257
  | Claude-4.6-Opus | 0.704 | 0.841 | 0.748 | 0.168 | 0.852 | 0.908 | 0.385 |
258
+ | DeepSeek V4.1 Flash | 0.695 | 0.794 | 0.770 | 0.207 | 0.851 | 0.851 | 0.285 |
259
  | Qwen3.6-35B-A3B | 0.694 | 0.859 | 0.784 | 0.099 | 0.822 | 0.904 | 0.232 |
260
  | Qwen3.6-27B | 0.691 | 0.863 | 0.768 | 0.071 | 0.833 | 0.919 | 0.195 |
261
  | Gemini-2.5-pro | 0.690 | 0.840 | 0.717 | 0.070 | 0.888 | 0.935 | 0.088 |
262
+ | Qwen3.8-27B | 0.688 | 0.800 | 0.758 | 0.236 | 0.812 | 0.836 | 0.391 |
263
  | Kimi K2 Instruct | 0.686 | 0.874 | 0.726 | 0.148 | 0.805 | 0.876 | 0.331 |
264
  | Cotype Light 3 | 0.685 | 0.855 | 0.770 | 0.125 | 0.873 | 0.799 | 0.318 |
265
  | GPT-5.4 | 0.683 | 0.764 | 0.627 | 0.351 | 0.826 | 0.846 | 0.432 |