To search datasets programmatically: GET https://api.databazaar.io/datasets?query=your-search

Full API docs: https://api.databazaar.io/llms.txt

Agent discovery: https://databazaar.io/.well-known/agent.json

Browse Data

1–24 of 32
Filters
Newest
images
$25

Shah Ali Azher

7 july 2025 · Updated Oct 26, 2026

This Data is for retouch purpose where people practice for retouching on adobe photoshop

Format
JPG
Downloads
0
Sample structureNot assessed
images
$25

Sanni Antikainen Raws

7 july 2025 · Updated Oct 26, 2026

This Data is for retouch purpose where people practice for retouching on adobe photoshop

Downloads
0
Sample structureNot assessed
images
Free

Conceptual Captions Image–Caption Pairs

Web images paired with natural-language captions harvested from alt-text descriptions, designed for training and evaluating image-captioning and vision-language models.

Records
5,341,263
Format
PARQUET
Downloads
2
Sample structure100/100
images
Free

Vessel Detection Labeled Satellite Image Patches

Validated satellite image patches labeled for maritime vessel detection, with object-detection and classification metadata. Exported from a boat-detection review workflow.

Records
1,920
Format
PARQUET
Downloads
2
Sample structure100/100
images
Free

Global Wheat Full Semantic Organ Segmentation

Labelled images for semantic segmentation of wheat plant organs (canopy, leaves, stems, heads) across diverse field conditions. Dataset spans multiple growing regions and environmental settings.

Records
1,096
Format
PARQUET
Downloads
5
Sample structure100/100
images
Free

RxRx3-core Phenomics Microscopy Image Challenge

Labeled microscopy images of cellular phenotypes from 735 genetic knockouts and 1,674 small-molecule perturbations, designed as a benchmark for learning cellular image representations.

Records
1,335,606
Format
PARQUET
Downloads
2
Sample structure100/100
images
Free

Pexels Synthetic Captions — InternVL2-40B

English captions for Pexels photos generated with InternVL2-40B-AWQ and grounded with original tags. Suitable for text-to-image and image-to-text model training.

Records
567,573
Format
PARQUET
Downloads
5
Sample structure96.8/100
images
Free

American Sign Language Video Dataset

Video dataset of ASL gestures across 2,208 distinct words, with preprocessed frames suitable for gesture recognition and sign language research.

Records
47
Format
PARQUET
Downloads
2
Sample structure100/100
images
Free

ShotBench Cinematic Understanding Benchmark

Expert-level question-answer pairs evaluating cinematic understanding in vision-language models, drawn from Oscar-nominated cinematography films across images and video clips.

Records
3,572
Format
PARQUET
Downloads
2
Sample structure100/100
images
Free

UAVIT-1M UAV Visual Instruction Tuning Dataset

Instruction-tuning dataset for low-altitude unmanned aerial vehicle visual understanding, covering 11 image- and region-level tasks across 1.24M samples.

Records
1,240,666
Format
PARQUET
Downloads
5
Sample structure100/100
images
Free

TextCaps Image Captioning Benchmark

Image captioning benchmark requiring optical character recognition and text reading within images. Formatted for multimodal model evaluation.

Records
28,408
Format
PARQUET
Downloads
4
Sample structure100/100
images
Free

Vero-600K Multi-Task Visual Reasoning RL Dataset

Reinforcement learning samples spanning 59 datasets across 6 visual reasoning categories, designed for training and evaluating vision-language models.

Records
606,000
Format
PARQUET
Downloads
3
Sample structure100/100
images
Free

PubTabNet Table Images with HTML Annotations

Scientific table images extracted from PubMed Central Open Access articles, paired with HTML structure annotations. Standard benchmark for image-based table recognition and document AI.

Records
223,100
Format
PARQUET
Downloads
5
Sample structure100/100
images
Free

EO-Data-1.5M Interleaved Vision-Text-Action Dataset

Multimodal dataset pairing visual observations, text descriptions, and robot actions to support embodied AI and robot learning tasks. Emphasizes temporal dynamics and causal relationships across modalities.

Records
1,422,808
Format
PARQUET
Downloads
2
Sample structure100/100
images
Free

FLUX-Reason-6M Bilingual Text-to-Image Dataset

Images synthesized by FLUX.1-dev paired with English and Chinese descriptions, designed to develop complex reasoning capabilities in text-to-image generative models.

Records
5,890,279
Format
PARQUET
Downloads
1
Sample structure90.6/100
images
Free

Text-to-Image Curated Pairs

Text-image pairs in webdataset format designed for fine-tuning text-to-image generative models.

Records
10,000
Format
PARQUET
Downloads
3
Sample structure100/100
images
Free

TreeOfLife-10M Biological Organism Images

Aggregated dataset of biological organism images paired with taxonomic labels across 454,000 taxa, compiled from iNaturalist, BIOSCAN-1M, and Encyclopedia of Life. Used to train vision-language models including BioCLIP.

Records
15,100
Format
PARQUET
Downloads
2
Sample structure100/100
images
Free

MMFineReason-SFT Multimodal Reasoning Samples

123,603 multimodal reasoning samples with Qwen3-VL-235B chain-of-thought traces, curated as the hardest 7% subset where smaller models consistently fail. Includes image and text data in parquet format.

Records
122,603
Format
PARQUET
Downloads
4
Sample structure94.2/100
images
Free

WorldCuisines VQA Multilingual Visual QA Benchmark

Multilingual visual question-answering dataset on global cuisines spanning 30 languages, with paired images and text questions/answers for vision-language model evaluation and fine-tuning.

Records
1,152,000
Format
PARQUET
Downloads
3
Sample structure100/100
images
Free

TreeOfLife-200M Biodiversity Image Dataset

Biodiversity images spanning 952,257 taxa from GBIF, EOL, BIOSCAN-5M, and FathomNet. Designed for computer vision and zero-shot taxonomic classification tasks.

Records
213,937,319
Format
PARQUET
Downloads
4
Sample structure98.5/100
images
Free

Nemotron VLM Dataset v2

Vision-language training dataset covering visual question answering, image-text-to-text, video-text-to-text, and document understanding tasks.

Records
4,518,886
Format
PARQUET
Downloads
3
Sample structure100/100
images
Free

FineVision Vision-Language Training Corpus

Image-text dataset with 17.3M images and 88.9M conversational turns. Designed for fine-tuning vision-language models with multi-subset organization.

Records
24,209,105
Format
PARQUET
Downloads
2
Sample structure100/100
images
Free

ScaleEdit-12M Image Editing Dataset

Synthetic instruction–image pairs across 23 editing task families, generated through the ScaleEditor framework. Automated verification may leave residual noise and model biases; instructions are English-only.

Records
12,369,418
Format
PARQUET
Downloads
2
Sample structure100/100
images
Free

Llama-Nemotron Vision-Language Model Dataset

Multimodal dataset for vision-language model training covering visual question answering, optical character recognition, image captioning, and image-to-text tasks. Includes approximately 2.9 million samples.

Records
2,863,834
Format
PARQUET
Downloads
2
Sample structure87.5/100