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
25–30 of 30MINT-1T ArXiv — Multimodal Interleaved ArXiv Papers (Text + Images)
ArXiv subset of MINT-1T: multimodal interleaved text-and-image documents extracted from ArXiv papers, designed for multimodal pretraining at scale. CC-BY-4.0.
MMFineReason 1.8M — Multimodal Reasoning Traces (Qwen3-VL-235B-Thinking)
1.8M multimodal reasoning samples with 5.1B solution tokens distilled from Qwen3-VL-235B-A22B-Thinking. Image+text inputs with detailed chain-of-thought annotations for math, science, and STEM visual reasoning. Apache-2.0.
LLaVA-OneVision-1.5 Instruction Data (22M Multimodal SFT)
22M curated instruction-tuning examples (image+text) used to train LLaVA-OneVision-1.5 LMMs. Apache-2.0, covers VQA, captioning, OCR, reasoning, and more.
Open-Access Museum Artwork Metadata — 10,000+ Works (1000–2025)
A consolidated metadata catalog of 10,000+ artworks from the worlds leading open-access museum collections. Each record includes artwork title, artist, creation date, medium, dimensions, department, culture/origin, classification, and direct image URLs. Sourced from the Metropolitan Museum of Art, Art Institute of Chicago, Cleveland Museum of Art, and Rijksmuseum open-access APIs. Ideal for training image classification models, art historical analysis, cultural heritage research, and recommendation systems. All records are normalized to a common schema with consistent field naming and formatting.
Historical Photograph Archive Metadata — 30 Archives, 6 Continents (1840–2024)
Consolidated metadata catalog of 15,000 historical photographs from 30 major archives across 6 continents, spanning nearly two centuries of photographic history from the daguerreotype era to the digital age. ## Sources - **Library of Congress**: American historical photography, Civil War documentation, FSA/OWI collection - **National Archives UK / Imperial War Museum**: British colonial and wartime photography - **Bibliothèque nationale de France / Musée d'Orsay**: French photographic heritage, early pictorialism - **Getty Research Institute**: Fine art and documentary photography - **Smithsonian Institution**: American cultural and scientific photography - **Bundesarchiv**: German historical and press photography - **National Diet Library / National Archives of Japan**: East Asian photographic records - **George Eastman Museum**: History of photography collection - **Rijksmuseum / Victoria and Albert Museum**: European art photography - **20+ additional national archives**: Australia, India, Brazil, Russia, South Africa, Sweden, and more ## Key Features - **30 columns** covering provenance, physical properties, digitization status, and art-historical context - **15 photographic media types**: daguerreotype, calotype, tintype, albumen print, gelatin silver, platinum, cyanotype, autochrome, chromogenic, Polaroid, digital capture, and more - **45 subject classifications**: portrait, landscape, war/conflict, documentary, scientific, botanical, street scene, and others - **Era-accurate media distribution**: daguerreotypes concentrated in 1840–1860, gelatin silver prints dominating 1885–1970, digital capture rising post-2000 - **Digitization metadata**: scan resolution (DPI), pixel dimensions, file sizes for 90%+ of records - **Condition assessments**: 6-level scale from excellent to deteriorated, with age-correlated degradation - **Rights status**: public domain flags for pre-1930 works, Creative Commons, restricted, and orphan work classifications - **Significance scoring**: 0–10 scale based on historical importance, with exhibition count tracking ## Use Cases - Digital humanities research on photographic history and visual culture - Archive digitization planning and prioritization - Photography market analysis and collection valuation - Machine learning training data for historical photo dating and classification - Cultural heritage preservation studies - Media history curriculum development
Museum Artwork & Image Metadata — 40 Museums, 6 Continents (1400–2024)
In-depth metadata dataset of 12,500 artworks and images from 40 major museums across 19 countries and 6 continents, spanning over 600 years of art history. ## Sources - **Metropolitan Museum of Art Open Access**: Artwork metadata, classifications, and provenance data - **Rijksmuseum API**: Dutch and European art collection records - **Art Institute of Chicago API**: American and European artwork metadata - **Smithsonian Open Access**: Photography and mixed media collections - **Europeana Collections**: Cross-institutional European art records - **Museum APIs worldwide**: Tokyo National Museum, National Museum of China, National Gallery (London), Louvre, and 30+ additional institutions ## Key Features - **28 columns** including artist demographics, physical dimensions, digital image specs, and art-historical classification - **8 classification types**: painting (35%), photograph (20%), print (15%), drawing (12%), sculpture (8%), watercolor (5%), mixed media (3%), digital art (2%) - **30 art movements** from Renaissance to Contemporary, with period-accurate movement assignments - **Image metadata**: pixel dimensions, file sizes, dominant color analysis (hex codes), and aspect ratios - **Public domain flags**: works created before 1929 flagged for open use - **Museum-level detail**: accession numbers, departments, city/country/continent geography ## Use Cases - Art market analysis and valuation modeling - Museum collection diversity studies - Computer vision training set curation (using image metadata to filter appropriate records) - Cultural heritage digitization planning - Art history trend analysis across periods and movements - Geographic distribution of global art collections