Skip to content

Datasets

Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,788 in the catalog so far, with new ones added every hour.

Data type

For which models

Find a dataset for your goal

Answered by Gemini 4 · the chat is kept only in this tab

DatasetBAAI/DenseFusion-1MTop pick

1M Dataset with Detailed Image Descriptions

A large multimodal set of 1 million to 10 million image-text pairs in English.

  • Question answering
  • Visual question answering
  • English
3K 441M–10M rows
DatasetHuggingFaceM4/DoclingMatix

Multimodal document recognition dataset

From 1 to 10 million document image-text pairs in English for training VLM models.

  • Visual question answering
  • Image text → text
  • English
2.9K 581M–10M rows
Datasetaipracticecafe/curated-danbooru-2026

Curated Danbooru anime illustrations with tags

Curated dataset of 330 thousand high-quality anime arts with tags and prompts for training image generators.

  • Text → image
  • Image → text
  • English
2.9K 17100K–1M rows
DatasetMMInstruction/ArxivQA

Multimodal QA on arXiv figures and charts

Over 100k multiple-choice questions based on scientific images from arXiv articles.

  • Image → text
  • English
2.8K 38100K–1M rows
Datasetopenbmb/RLAIF-V-Dataset

RLAIF-V data for trustworthy VLMs

English images and text for training and validating trustworthy multimodal models via RLAIF.

  • Image text → text
  • Visual question answering
  • Multimodal
  • English
2.6K 22710K–100K rows
DatasetOpenDataArena/MMFineReason-1.8M-Qwen3-VL-235B-ThinkingTop pick

Multimodal reasoning on 1.8M examples

1.8M multimodal examples with detailed reasoning distilled from Qwen3-VL-235B-Thinking.

  • Visual question answering
  • Question answering
  • Text generation
  • English
2.6K 1271M–10M rows
DatasetHarland/OmniVChatTop pick

Native audiovisual dialog dataset

Data from Alibaba and Chinese universities for training and evaluating models processing simultaneous video and voice streams.

  • Video text → text
  • English
  • Chinese
2.6K 581K–10K rows
Datasetgoogle/MapTrace

Synthetic map path tracing

2M synthetic examples for map QA and path tracing in English.

  • Image → text
  • Visual question answering
  • English
2.5K 12110K–100K rows
DatasetPRAIG/SMB

Sheet music recognition (OMR) benchmark

Printed sheet music images with region annotations for testing optical music recognition systems.

  • Image → text
  • Image segmentation
  • Text retrieval
2.5K 17up to 1K rows🔒 on request
Datasetnvidia/Nemotron-Image-Training-v3.5Top pick

Nemotron VL training data

Collection of multimodal training annotations for the Nemotron Super 3.5 VL model.

  • Visual question answering
  • Image text → text
  • English
2.4K 81M–10M rows
Datasetjovianzm/Pexels-400k

400k Pexels Videos with Captions and Metadata

400,476 videos with previews, view counts, explicit flags, and captions for video-text models.

  • Image → text
  • Text → image
  • Text → video
  • English
2.4K 70100K–1M rows
Datasetfacebook/winoground

Winoground: compositional reasoning

Image and caption pairs with identical words in different orders to test vision-language models.

  • Image → text
  • Text → image
  • Image classification
  • English
2.3K 124up to 1K rows🔒 on request
DatasetInfi-MM/InfiMM-WebMath-40BTop pick

InfiMM-WebMath-40B multimodal math corpus

Large multimodal dataset of math problems with text and images in English and Chinese.

  • Image text → text
  • English
  • Chinese
2.3K 6910M–100M rows
Datasetmoonshotai/WorldVQATop pick

Visual World Knowledge Benchmark for MLLMs

Image-based questions to evaluate atomic world knowledge in multimodal models.

  • Visual question answering
  • English
  • Chinese
2.3K 671K–10K rows
Datasetshunk031/STAIR-Captions

Large Japanese image captioning dataset

A large-scale STAIR Captions dataset with 820,310 Japanese image captions for text-to-image and image-to-text generation tasks.

  • Image → text
  • Japanese
2.3K 7100K–1M rows
DatasetCareerflow/ResumeBench

Benchmark for resume parsing

Benchmark for evaluating the accuracy of structured data extraction from resumes by JSON Schema.

  • Document question answering
  • Embeddings
  • English
2.2K 3up to 1K rows🔒 on request
DatasetLin-Chen/ShareGPT4V

1.2M image captions by GPT-4V

Multimodal English captions for image-text alignment and training captioning models.

  • Visual question answering
  • Question answering
  • English
2.2K 3281M–10M rows
DatasetByteDance/MTVQA

Multilingual Text-Image VQA (9 Languages)

Image-text question-answer pairs in nine languages, including Russian, for evaluating and training multimodal models.

  • Visual question answering
  • Image → text
  • Multilingual
  • Arabic
  • German
  • +7
2.2K 421K–10K rows
Datasetbevaya/pubmed-ocr

PubMed-OCR: scientific PDF layout markup

PubMed Central page OCR markup with bounding boxes for layout-aware tasks.

  • Image → text
  • Image text → text
  • English
2.1K 731M–10M rows
DatasetRekaAI/VibeEval

Vibe-Eval benchmark for multimodal chat models

A compact set of challenging vision-language tasks with reference answers in four languages.

  • Image → text
  • Image classification
  • English
  • Polish
  • Chinese
  • +1
1.9K 49up to 1K rows