Benchmark Almanac

UCF101

QuanSun/EVA02-CLIP-bigE-14-plus 93.3% tops its leaderboard
Not in model cards

Models recognize one of 101 human actions in video clips.

MTEB reports accuracy on its UCF101 task.

Released
December 2012
Kind
Video understanding
Level
Everyday
Scored
Checked automatically
Answers
Prediction
Tasks
All public
Measures
Capability
Language
English
In and out
Video → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
In views
Vision
49
Entries on MTEB’s leaderboard, 0 models we track

QuanSun/EVA02-CLIP-bigE-14-plus tops the leaderboard with 93.3%

Best score per entry from MTEB’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.

49 entries. Scroll sideways for the rest.

  1. 93.3 QuanSun/EVA02-CLIP-bigE-14-plus
  2. 92.1 QuanSun/EVA02-CLIP-bigE-14
  3. 91.7 google/siglip-so400m-patch14-384
  4. 90.8 laion/CLIP-ViT-bigG-14-laion2B-39B-b160k
  5. 90.5 google/siglip-large-patch16-384
  6. 89.8 laion/CLIP-ViT-H-14-laion2B-s32B-b79K
  7. 89.2 facebook/dinov2-giant
  8. 89.1 laion/CLIP-ViT-g-14-laion2B-s34B-b88K
  9. 88.9 google/siglip-large-patch16-256
  10. 88.6 laion/CLIP-ViT-L-14-DataComp.XL-s13B-b90K
  11. 88.4 dnotitia/DNA-VL-STEER-2B
  12. 88.3 Salesforce/blip2-opt-2.7b
  13. 88.1 facebook/dinov2-large
  14. 87.8 openai/clip-vit-large-patch14
  15. 87.7 google/siglip-base-patch16-512
  16. 87.5 Salesforce/blip2-opt-6.7b-coco
  17. 87.5 laion/CLIP-ViT-L-14-laion2B-s32B-b82K
  18. 87.0 google/siglip-base-patch16-384
  19. 86.6 jinaai/jina-embeddings-v5-omni-small
  20. 86.1 royokong/e5-v
  21. 85.5 facebook/dinov2-base
  22. 85.2 google/siglip-base-patch16-256
  23. 84.8 google/siglip-base-patch16-224
  24. 84.5 Salesforce/blip-itm-large-coco
  25. 84.3 mongodb/voyage-multimodal-3
  26. 83.8 google/siglip-base-patch16-256-multilingual
  27. 82.8 laion/CLIP-ViT-B-16-DataComp.XL-s13B-b90K
  28. 82.2 Salesforce/blip-itm-large-flickr
  29. 81.9 openai/clip-vit-base-patch16
  30. 81.9 QuanSun/EVA02-CLIP-L-14
  31. 81.8 Salesforce/blip-itm-base-coco
  32. 81.1 laion/CLIP-ViT-B-32-laion2B-s34B-b79K
  33. 80.6 Salesforce/blip-image-captioning-large
  34. 80.5 facebook/dinov2-small
  35. 80.1 Salesforce/blip-itm-base-flickr
  36. 80.1 kakaobrain/align-base
  37. 78.9 laion/CLIP-ViT-B-32-DataComp.XL-s13B-b90K
  38. 78.8 openai/clip-vit-base-patch32
  39. 78.7 jinaai/jina-embeddings-v5-omni-nano
  40. 77.8 BAAI/bge-visualized-base
  41. 76.7 jinaai/jina-clip-v1
  42. 76.5 TIGER-Lab/VLM2Vec-LoRA
  43. 76.5 TIGER-Lab/VLM2Vec-Full
  44. 76.2 BAAI/bge-visualized-m3
  45. 73.5 nomic-ai/nomic-embed-vision-v1.5
  46. 72.9 QuanSun/EVA02-CLIP-B-16
  47. 72.9 Salesforce/blip-image-captioning-base
  48. 71.0 nyu-visionx/moco-v3-vit-l
  49. 70.1 nyu-visionx/moco-v3-vit-b
  1. 1 QuanSun/EVA02-CLIP-bigE-14-plus 93.3%
  2. 2 QuanSun/EVA02-CLIP-bigE-14 92.1%
  3. 3 google/siglip-so400m-patch14-384 91.7%
  4. 4 laion/CLIP-ViT-bigG-14-laion2B-39B-b160k 90.8%
  5. 5 google/siglip-large-patch16-384 90.5%
  6. 6 laion/CLIP-ViT-H-14-laion2B-s32B-b79K 89.8%
  7. 7 facebook/dinov2-giant 89.2%
  8. 8 laion/CLIP-ViT-g-14-laion2B-s34B-b88K 89.1%
  9. 9 google/siglip-large-patch16-256 88.9%
  10. 10 laion/CLIP-ViT-L-14-DataComp.XL-s13B-b90K 88.6%
  11. 11 dnotitia/DNA-VL-STEER-2B 88.4%
  12. 12 Salesforce/blip2-opt-2.7b 88.3%
  13. 13 facebook/dinov2-large 88.1%
  14. 14 openai/clip-vit-large-patch14 87.8%
  15. 15 google/siglip-base-patch16-512 87.7%
  16. 16 Salesforce/blip2-opt-6.7b-coco 87.5%
  17. 17 laion/CLIP-ViT-L-14-laion2B-s32B-b82K 87.5%
  18. 18 google/siglip-base-patch16-384 87%
  19. 19 jinaai/jina-embeddings-v5-omni-small 86.6%
  20. 20 royokong/e5-v 86.1%
  21. 21 facebook/dinov2-base 85.5%
  22. 22 google/siglip-base-patch16-256 85.2%
  23. 23 google/siglip-base-patch16-224 84.8%
  24. 24 Salesforce/blip-itm-large-coco 84.5%
  25. 25 mongodb/voyage-multimodal-3 84.3%
  26. 26 google/siglip-base-patch16-256-multilingual 83.8%
  27. 27 laion/CLIP-ViT-B-16-DataComp.XL-s13B-b90K 82.8%
  28. 28 Salesforce/blip-itm-large-flickr 82.2%
  29. 29 openai/clip-vit-base-patch16 81.9%
  30. 30 QuanSun/EVA02-CLIP-L-14 81.9%
  31. 31 Salesforce/blip-itm-base-coco 81.8%
  32. 32 laion/CLIP-ViT-B-32-laion2B-s34B-b79K 81.1%
  33. 33 Salesforce/blip-image-captioning-large 80.6%
  34. 34 facebook/dinov2-small 80.5%
  35. 35 Salesforce/blip-itm-base-flickr 80.1%
  36. 36 kakaobrain/align-base 80.1%
  37. 37 laion/CLIP-ViT-B-32-DataComp.XL-s13B-b90K 78.9%
  38. 38 openai/clip-vit-base-patch32 78.8%
  39. 39 jinaai/jina-embeddings-v5-omni-nano 78.7%
  40. 40 BAAI/bge-visualized-base 77.8%
  41. 41 jinaai/jina-clip-v1 76.7%
  42. 42 TIGER-Lab/VLM2Vec-LoRA 76.5%
  43. 43 TIGER-Lab/VLM2Vec-Full 76.5%
  44. 44 BAAI/bge-visualized-m3 76.2%
  45. 45 nomic-ai/nomic-embed-vision-v1.5 73.5%
  46. 46 QuanSun/EVA02-CLIP-B-16 72.9%
  47. 47 Salesforce/blip-image-captioning-base 72.9%
  48. 48 nyu-visionx/moco-v3-vit-l 71%
  49. 49 nyu-visionx/moco-v3-vit-b 70.1%

Source: MTEB’s leaderboard

#EntryScoreSettingDateLinkSource
1 QuanSun/EVA02-CLIP-bigE-14-plus open weights, 5B 93.3% huggingface.co Official leaderboard
2 QuanSun/EVA02-CLIP-bigE-14 open weights, 4.7B 92.1% huggingface.co Official leaderboard
3 google/siglip-so400m-patch14-384 open weights, 0.9B 91.7% huggingface.co Official leaderboard
4 laion/CLIP-ViT-bigG-14-laion2B-39B-b160k open weights, 2.5B 90.8% huggingface.co Official leaderboard
5 google/siglip-large-patch16-384 open weights, 0.7B 90.5% huggingface.co Official leaderboard
6 laion/CLIP-ViT-H-14-laion2B-s32B-b79K open weights, 1B 89.8% huggingface.co Official leaderboard
7 facebook/dinov2-giant open weights, 1.1B 89.2% huggingface.co Official leaderboard
8 laion/CLIP-ViT-g-14-laion2B-s34B-b88K open weights, 1.4B 89.1% huggingface.co Official leaderboard
9 google/siglip-large-patch16-256 open weights, 0.7B 88.9% huggingface.co Official leaderboard
10 laion/CLIP-ViT-L-14-DataComp.XL-s13B-b90K open weights, 0.4B 88.6% huggingface.co Official leaderboard
11 dnotitia/DNA-VL-STEER-2B open weights, 2.1B 88.4% huggingface.co Official leaderboard
12 Salesforce/blip2-opt-2.7b open weights, 3.7B 88.3% huggingface.co Official leaderboard
13 facebook/dinov2-large open weights, 0.3B 88.1% huggingface.co Official leaderboard
14 openai/clip-vit-large-patch14 open weights, 0.4B 87.8% huggingface.co Official leaderboard
15 google/siglip-base-patch16-512 open weights, 0.2B 87.7% huggingface.co Official leaderboard
16 Salesforce/blip2-opt-6.7b-coco open weights, 7.8B 87.5% huggingface.co Official leaderboard
17 laion/CLIP-ViT-L-14-laion2B-s32B-b82K open weights, 0.4B 87.5% huggingface.co Official leaderboard
18 google/siglip-base-patch16-384 open weights, 0.2B 87% huggingface.co Official leaderboard
19 jinaai/jina-embeddings-v5-omni-small open weights, 1.6B 86.6% huggingface.co Official leaderboard
20 royokong/e5-v open weights, 8.4B 86.1% huggingface.co Official leaderboard
21 facebook/dinov2-base open weights, 0.1B 85.5% huggingface.co Official leaderboard
22 google/siglip-base-patch16-256 open weights, 0.2B 85.2% huggingface.co Official leaderboard
23 google/siglip-base-patch16-224 open weights, 0.2B 84.8% huggingface.co Official leaderboard
24 Salesforce/blip-itm-large-coco open weights, 0.4B 84.5% huggingface.co Official leaderboard
25 mongodb/voyage-multimodal-3 closed weights 84.3% huggingface.co Official leaderboard
26 google/siglip-base-patch16-256-multilingual open weights, 0.4B 83.8% huggingface.co Official leaderboard
27 laion/CLIP-ViT-B-16-DataComp.XL-s13B-b90K open weights, 0.2B 82.8% huggingface.co Official leaderboard
28 Salesforce/blip-itm-large-flickr open weights, 0.4B 82.2% huggingface.co Official leaderboard
29 openai/clip-vit-base-patch16 open weights, 0.2B 81.9% huggingface.co Official leaderboard
30 QuanSun/EVA02-CLIP-L-14 open weights, 0.4B 81.9% huggingface.co Official leaderboard
31 Salesforce/blip-itm-base-coco open weights, 0.2B 81.8% huggingface.co Official leaderboard
32 laion/CLIP-ViT-B-32-laion2B-s34B-b79K open weights, 0.2B 81.1% huggingface.co Official leaderboard
33 Salesforce/blip-image-captioning-large open weights, 0.4B 80.6% huggingface.co Official leaderboard
34 facebook/dinov2-small open weights, 0B 80.5% huggingface.co Official leaderboard
35 Salesforce/blip-itm-base-flickr open weights, 0.2B 80.1% huggingface.co Official leaderboard
36 kakaobrain/align-base open weights, 0.2B 80.1% huggingface.co Official leaderboard
37 laion/CLIP-ViT-B-32-DataComp.XL-s13B-b90K open weights, 0.2B 78.9% huggingface.co Official leaderboard
38 openai/clip-vit-base-patch32 open weights, 0.2B 78.8% huggingface.co Official leaderboard
39 jinaai/jina-embeddings-v5-omni-nano open weights, 1B 78.7% huggingface.co Official leaderboard
40 BAAI/bge-visualized-base open weights, 0.2B 77.8% huggingface.co Official leaderboard
41 jinaai/jina-clip-v1 open weights, 0.2B 76.7% huggingface.co Official leaderboard
42 TIGER-Lab/VLM2Vec-LoRA open weights, 4.2B 76.5% huggingface.co Official leaderboard
43 TIGER-Lab/VLM2Vec-Full open weights, 4.1B 76.5% huggingface.co Official leaderboard
44 BAAI/bge-visualized-m3 open weights, 0.9B 76.2% huggingface.co Official leaderboard
45 nomic-ai/nomic-embed-vision-v1.5 open weights, 0.1B 73.5% huggingface.co Official leaderboard
46 QuanSun/EVA02-CLIP-B-16 open weights, 0.1B 72.9% huggingface.co Official leaderboard
47 Salesforce/blip-image-captioning-base open weights, 0.2B 72.9% huggingface.co Official leaderboard
48 nyu-visionx/moco-v3-vit-l open weights, 0.3B 71% github.com Official leaderboard
49 nyu-visionx/moco-v3-vit-b open weights, 0.1B 70.1% github.com 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 Audio and video

All 695 in Audio and video
  • Video-MMMU In use 300 expert-level lecture videos with questions testing perception, comprehension and adaptation of knowledge. Accuracy. Qwen 3.8 Max 88.7%
  • LVBench In use Hours-long video understanding with multiple-choice questions on reasoning, grounding and summarization. Gemini 4 Argon 91.7%
  • MMVU In use 3,000 expert-annotated questions on videos across 27 disciplines. Multiple-choice and open-ended accuracy. Qwen 3.8 Max 82.4%
  • MLVU In use Multi-task long video understanding (minutes to hours) with multiple-choice and generation tasks. Accuracy. Qwen 3.8 Max 90.8%
  • FLEURS In use Models transcribe read speech in 102 languages built from FLoRes-101 sentences. Model cards usually report word error rate, so lower scores are better. Phi-4-multimodal 4.00%
  • VoiceBench In use Spoken instructions with speaker, environment and content variation. Voice assistant responses scored by GPT judge or rules. Qwen3.5-Omni-Plus 93.1%

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 (49 entries).
  • Pass Models scored 49 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: 49 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