Benchmark Almanac

Stanford Online Products

vits16_sop 86.6% tops its leaderboard
Not in model cards

Models retrieve photographs of the same product. The test compares learned image features on products outside the training set.

Made by
Hyun Oh Song and colleagues
Released
November 2015
Kind
General visual questions
Level
Everyday
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Measures
Capability
Language
English
In and out
Images → Text
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
In views
Vision
23
Entries on Open Metric Learning’s leaderboard, 0 models we track

vits16_sop tops the leaderboard with 86.6%

Best score per entry from Open Metric Learning’s leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

  1. 1 vits16_sop 224 x 224 86.6%
  2. 2 vitl14_336px_unicom 224 x 224 74.5%
  3. 3 vitl14_unicom 224 x 224 72.6%
  4. 4 vitb16_unicom 224 x 224 70%
  5. 5 vitb32_unicom 224 x 224 69%
  6. 6 vitb8_dino 224 x 224 68.9%
  7. 7 openai_vitl14_224 224 x 224 67%
  8. 8 vitb16_dino 224 x 224 65.8%
  9. 9 vits8_dino 224 x 224 65.1%
  10. 10 openai_vitb16_224 224 x 224 64.8%
  11. 11 vits16_dino 224 x 224 64.8%
  12. 12 openai_vitb32_224 224 x 224 61.2%
  13. 13 vitl14_dinov2 224 x 224 57.6%
  14. 14 vitl14_reg_dinov2 224 x 224 57.1%
  15. 15 vits14_dinov2 224 x 224 56.6%
  16. 16 vits14_reg_dinov2 224 x 224 56.6%
  17. 17 sber_vitb16_224 224 x 224 56.5%
  18. 18 vitb14_dinov2 224 x 224 56.5%
  19. 19 vitb14_reg_dinov2 224 x 224 55.7%
  20. 20 sber_vitb32_224 224 x 224 54.7%
  21. 21 resnet50_imagenet1k_v1 224 x 224 51.5%
  22. 22 sber_vitl14_224 224 x 224 51.2%
  23. 23 resnet50_moco_v2 224 x 224 49.3%

Source: Open Metric Learning’s leaderboard

#EntryScoreSettingDateSource
1 vits16_sop 86.6% 224 x 224 Official leaderboard
2 vitl14_336px_unicom 74.5% 224 x 224 Official leaderboard
3 vitl14_unicom 72.6% 224 x 224 Official leaderboard
4 vitb16_unicom 70% 224 x 224 Official leaderboard
5 vitb32_unicom 69% 224 x 224 Official leaderboard
6 vitb8_dino 68.9% 224 x 224 Official leaderboard
7 openai_vitl14_224 67% 224 x 224 Official leaderboard
8 vitb16_dino 65.8% 224 x 224 Official leaderboard
9 vits8_dino 65.1% 224 x 224 Official leaderboard
10 openai_vitb16_224 64.8% 224 x 224 Official leaderboard
11 vits16_dino 64.8% 224 x 224 Official leaderboard
12 openai_vitb32_224 61.2% 224 x 224 Official leaderboard
13 vitl14_dinov2 57.6% 224 x 224 Official leaderboard
14 vitl14_reg_dinov2 57.1% 224 x 224 Official leaderboard
15 vits14_dinov2 56.6% 224 x 224 Official leaderboard
16 vits14_reg_dinov2 56.6% 224 x 224 Official leaderboard
17 sber_vitb16_224 56.5% 224 x 224 Official leaderboard
18 vitb14_dinov2 56.5% 224 x 224 Official leaderboard
19 vitb14_reg_dinov2 55.7% 224 x 224 Official leaderboard
20 sber_vitb32_224 54.7% 224 x 224 Official leaderboard
21 resnet50_imagenet1k_v1 51.5% 224 x 224 Official leaderboard
22 sber_vitl14_224 51.2% 224 x 224 Official leaderboard
23 resnet50_moco_v2 49.3% 224 x 224 Official leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Vision

All 636 in Vision
  • MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy. Qwen 3.6 Plus 86%
  • MMMU-Pro In use Harder MMMU: college-level multimodal questions with more options and vision-only inputs to remove text shortcuts. GPT-5.6 Sol 84.6%
  • CharXiv In use Chart understanding on real arXiv figures. Reasoning questions require synthesizing information from complex charts. Claude Fable 5 93.5%
  • MathVista In use Math reasoning in visual contexts (charts, geometry, puzzles) from 31 datasets. Answer accuracy. Seed2.1 Pro 90.7%
  • MATH-Vision (MathVision) In use 3,040 competition math problems with visual context. Answer accuracy. Qwen 3.8 Max 97.7%
  • MMStar In use Models answer 1,500 human-curated image questions that cannot be solved without the image, across 6 core capabilities. Scored by accuracy. Qwen 3.8 Max 85.9%

Data audit

All 8 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
  • Pass Independent runs Scores from the official leaderboard (23 entries).
  • Pass Models scored 23 entries on its leaderboard, none a model we track.
  • Note Newest score No scores yet.
  • Note Frontier models No models scored yet.
  • Note Headroom No scores yet.
Leaderboard import 3 pass
  • Pass Importer We read the board on Open Metric Learning: 23 entries at the last read, none a model we track.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check