Benchmark Almanac

Fashion IQ

mongodb/voyage-multimodal-3 15.5% tops its leaderboard
Not in model cards

Models retrieve a garment using a reference photograph and text describing how the target garment differs.

The task uses comparative descriptions supplied by people.

Made by
Hui Wu and colleagues
Released
May 2019
Kind
General visual questions
Level
Everyday
Scored
Checked automatically
Answers
Prediction
Tasks
All public
Measures
Capability
Language
English
In and out
Images, Text → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
In views
Vision
44
Entries on MTEB’s leaderboard, 0 models we track

mongodb/voyage-multimodal-3 tops the leaderboard with 15.5%

Best score per entry from MTEB’s leaderboard, read on November 20th, 2026. MTEB reports test-set nDCG at 10, shown as a percentage. 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.

44 entries. Scroll sideways for the rest.

  1. 15.5 mongodb/voyage-multimodal-3
  2. 11.1 google/siglip-so400m-patch14-384
  3. 10.9 QuanSun/EVA02-CLIP-bigE-14-plus
  4. 9.92 QuanSun/EVA02-CLIP-bigE-14
  5. 8.84 dnotitia/DNA-VL-STEER-2B
  6. 7.64 google/siglip-large-patch16-384
  7. 7.56 QuanSun/EVA02-CLIP-L-14
  8. 6.86 google/siglip-base-patch16-512
  9. 6.75 google/siglip-large-patch16-256
  10. 6.73 kakaobrain/align-base
  11. 6.65 Salesforce/blip-itm-large-coco
  12. 6.59 Salesforce/blip-itm-large-flickr
  13. 6.49 google/siglip-base-patch16-384
  14. 6.12 google/siglip-base-patch16-224
  15. 6.03 QuanSun/EVA02-CLIP-B-16
  16. 5.76 royokong/e5-v
  17. 5.53 google/siglip-base-patch16-256-multilingual
  18. 5.51 google/siglip-base-patch16-256
  19. 5.38 Salesforce/blip-itm-base-coco
  20. 5.31 BAAI/bge-visualized-m3
  21. 5.16 laion/CLIP-ViT-bigG-14-laion2B-39B-b160k
  22. 5.07 LanguageBind/LanguageBind_Omni
  23. 5.01 Salesforce/blip-itm-base-flickr
  24. 4.52 BAAI/bge-visualized-base
  25. 4.51 jinaai/jina-clip-v1
  26. 4.39 laion/CLIP-ViT-L-14-DataComp.XL-s13B-b90K
  27. 4.19 laion/CLIP-ViT-H-14-laion2B-s32B-b79K
  28. 4.13 laion/CLIP-ViT-L-14-laion2B-s32B-b82K
  29. 3.92 laion/CLIP-ViT-g-14-laion2B-s34B-b88K
  30. 3.44 laion/CLIP-ViT-B-16-DataComp.XL-s13B-b90K
  31. 3.11 openai/clip-vit-large-patch14
  32. 3.05 laion/CLIP-ViT-B-32-laion2B-s34B-b79K
  33. 2.83 laion/CLIP-ViT-B-32-DataComp.XL-s13B-b90K
  34. 2.59 openai/clip-vit-base-patch16
  35. 2.21 openai/clip-vit-base-patch32
  36. 2.21 TIGER-Lab/VLM2Vec-Full
  37. 2.12 TIGER-Lab/VLM2Vec-LoRA
  38. 2.02 nomic-ai/nomic-embed-vision-v1.5
  39. 1.86 Salesforce/blip2-opt-6.7b-coco
  40. 1.63 Salesforce/blip2-opt-2.7b
  41. 1.32 jinaai/jina-embeddings-v5-omni-nano
  42. 1.03 jinaai/jina-embeddings-v5-omni-small
  43. 0.73 Salesforce/blip-image-captioning-large
  44. 0.63 Salesforce/blip-image-captioning-base
  1. 1 mongodb/voyage-multimodal-3 15.5%
  2. 2 google/siglip-so400m-patch14-384 11.1%
  3. 3 QuanSun/EVA02-CLIP-bigE-14-plus 10.9%
  4. 4 QuanSun/EVA02-CLIP-bigE-14 9.92%
  5. 5 dnotitia/DNA-VL-STEER-2B 8.84%
  6. 6 google/siglip-large-patch16-384 7.64%
  7. 7 QuanSun/EVA02-CLIP-L-14 7.56%
  8. 8 google/siglip-base-patch16-512 6.86%
  9. 9 google/siglip-large-patch16-256 6.75%
  10. 10 kakaobrain/align-base 6.73%
  11. 11 Salesforce/blip-itm-large-coco 6.65%
  12. 12 Salesforce/blip-itm-large-flickr 6.59%
  13. 13 google/siglip-base-patch16-384 6.49%
  14. 14 google/siglip-base-patch16-224 6.12%
  15. 15 QuanSun/EVA02-CLIP-B-16 6.03%
  16. 16 royokong/e5-v 5.76%
  17. 17 google/siglip-base-patch16-256-multilingual 5.53%
  18. 18 google/siglip-base-patch16-256 5.51%
  19. 19 Salesforce/blip-itm-base-coco 5.38%
  20. 20 BAAI/bge-visualized-m3 5.31%
  21. 21 laion/CLIP-ViT-bigG-14-laion2B-39B-b160k 5.16%
  22. 22 LanguageBind/LanguageBind_Omni 5.07%
  23. 23 Salesforce/blip-itm-base-flickr 5.01%
  24. 24 BAAI/bge-visualized-base 4.52%
  25. 25 jinaai/jina-clip-v1 4.51%
  26. 26 laion/CLIP-ViT-L-14-DataComp.XL-s13B-b90K 4.39%
  27. 27 laion/CLIP-ViT-H-14-laion2B-s32B-b79K 4.19%
  28. 28 laion/CLIP-ViT-L-14-laion2B-s32B-b82K 4.13%
  29. 29 laion/CLIP-ViT-g-14-laion2B-s34B-b88K 3.92%
  30. 30 laion/CLIP-ViT-B-16-DataComp.XL-s13B-b90K 3.44%
  31. 31 openai/clip-vit-large-patch14 3.11%
  32. 32 laion/CLIP-ViT-B-32-laion2B-s34B-b79K 3.05%
  33. 33 laion/CLIP-ViT-B-32-DataComp.XL-s13B-b90K 2.83%
  34. 34 openai/clip-vit-base-patch16 2.59%
  35. 35 openai/clip-vit-base-patch32 2.21%
  36. 36 TIGER-Lab/VLM2Vec-Full 2.21%
  37. 37 TIGER-Lab/VLM2Vec-LoRA 2.12%
  38. 38 nomic-ai/nomic-embed-vision-v1.5 2.02%
  39. 39 Salesforce/blip2-opt-6.7b-coco 1.86%
  40. 40 Salesforce/blip2-opt-2.7b 1.63%
  41. 41 jinaai/jina-embeddings-v5-omni-nano 1.32%
  42. 42 jinaai/jina-embeddings-v5-omni-small 1.03%
  43. 43 Salesforce/blip-image-captioning-large 0.73%
  44. 44 Salesforce/blip-image-captioning-base 0.63%

Source: MTEB’s leaderboard

#EntryScoreSettingDateLinkSource
1 mongodb/voyage-multimodal-3 closed weights 15.5% huggingface.co Official leaderboard
2 google/siglip-so400m-patch14-384 open weights, 0.9B 11.1% huggingface.co Official leaderboard
3 QuanSun/EVA02-CLIP-bigE-14-plus open weights, 5B 10.9% huggingface.co Official leaderboard
4 QuanSun/EVA02-CLIP-bigE-14 open weights, 4.7B 9.92% huggingface.co Official leaderboard
5 dnotitia/DNA-VL-STEER-2B open weights, 2.1B 8.84% huggingface.co Official leaderboard
6 google/siglip-large-patch16-384 open weights, 0.7B 7.64% huggingface.co Official leaderboard
7 QuanSun/EVA02-CLIP-L-14 open weights, 0.4B 7.56% huggingface.co Official leaderboard
8 google/siglip-base-patch16-512 open weights, 0.2B 6.86% huggingface.co Official leaderboard
9 google/siglip-large-patch16-256 open weights, 0.7B 6.75% huggingface.co Official leaderboard
10 kakaobrain/align-base open weights, 0.2B 6.73% huggingface.co Official leaderboard
11 Salesforce/blip-itm-large-coco open weights, 0.4B 6.65% huggingface.co Official leaderboard
12 Salesforce/blip-itm-large-flickr open weights, 0.4B 6.59% huggingface.co Official leaderboard
13 google/siglip-base-patch16-384 open weights, 0.2B 6.49% huggingface.co Official leaderboard
14 google/siglip-base-patch16-224 open weights, 0.2B 6.12% huggingface.co Official leaderboard
15 QuanSun/EVA02-CLIP-B-16 open weights, 0.1B 6.03% huggingface.co Official leaderboard
16 royokong/e5-v open weights, 8.4B 5.76% huggingface.co Official leaderboard
17 google/siglip-base-patch16-256-multilingual open weights, 0.4B 5.53% huggingface.co Official leaderboard
18 google/siglip-base-patch16-256 open weights, 0.2B 5.51% huggingface.co Official leaderboard
19 Salesforce/blip-itm-base-coco open weights, 0.2B 5.38% huggingface.co Official leaderboard
20 BAAI/bge-visualized-m3 open weights, 0.9B 5.31% huggingface.co Official leaderboard
21 laion/CLIP-ViT-bigG-14-laion2B-39B-b160k open weights, 2.5B 5.16% huggingface.co Official leaderboard
22 LanguageBind/LanguageBind_Omni open weights, 1.2B 5.07% github.com Official leaderboard
23 Salesforce/blip-itm-base-flickr open weights, 0.2B 5.01% huggingface.co Official leaderboard
24 BAAI/bge-visualized-base open weights, 0.2B 4.52% huggingface.co Official leaderboard
25 jinaai/jina-clip-v1 open weights, 0.2B 4.51% huggingface.co Official leaderboard
26 laion/CLIP-ViT-L-14-DataComp.XL-s13B-b90K open weights, 0.4B 4.39% huggingface.co Official leaderboard
27 laion/CLIP-ViT-H-14-laion2B-s32B-b79K open weights, 1B 4.19% huggingface.co Official leaderboard
28 laion/CLIP-ViT-L-14-laion2B-s32B-b82K open weights, 0.4B 4.13% huggingface.co Official leaderboard
29 laion/CLIP-ViT-g-14-laion2B-s34B-b88K open weights, 1.4B 3.92% huggingface.co Official leaderboard
30 laion/CLIP-ViT-B-16-DataComp.XL-s13B-b90K open weights, 0.2B 3.44% huggingface.co Official leaderboard
31 openai/clip-vit-large-patch14 open weights, 0.4B 3.11% huggingface.co Official leaderboard
32 laion/CLIP-ViT-B-32-laion2B-s34B-b79K open weights, 0.2B 3.05% huggingface.co Official leaderboard
33 laion/CLIP-ViT-B-32-DataComp.XL-s13B-b90K open weights, 0.2B 2.83% huggingface.co Official leaderboard
34 openai/clip-vit-base-patch16 open weights, 0.2B 2.59% huggingface.co Official leaderboard
35 openai/clip-vit-base-patch32 open weights, 0.2B 2.21% huggingface.co Official leaderboard
36 TIGER-Lab/VLM2Vec-Full open weights, 4.1B 2.21% huggingface.co Official leaderboard
37 TIGER-Lab/VLM2Vec-LoRA open weights, 4.2B 2.12% huggingface.co Official leaderboard
38 nomic-ai/nomic-embed-vision-v1.5 open weights, 0.1B 2.02% huggingface.co Official leaderboard
39 Salesforce/blip2-opt-6.7b-coco open weights, 7.8B 1.86% huggingface.co Official leaderboard
40 Salesforce/blip2-opt-2.7b open weights, 3.7B 1.63% huggingface.co Official leaderboard
41 jinaai/jina-embeddings-v5-omni-nano open weights, 1B 1.32% huggingface.co Official leaderboard
42 jinaai/jina-embeddings-v5-omni-small open weights, 1.6B 1.03% huggingface.co Official leaderboard
43 Salesforce/blip-image-captioning-large open weights, 0.4B 0.73% huggingface.co Official leaderboard
44 Salesforce/blip-image-captioning-base open weights, 0.2B 0.63% huggingface.co 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 (44 entries).
  • Pass Models scored 44 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 MTEB: 44 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