datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
gpt-image-2-prompts-datasets
🖼️ GPT Image 2 Prompt Dataset
🖼️ The ultimate GPT Image 2 prompt dataset (5GB+). 15,000+ image generation prompts with full metadata and preview images. Truly open source: No login, no ads, no redirection. Just pure data for AI image creators.
This project is a massive collection of prompts used for OpenAI's GPT Image 2 model and the resulting generated images. The entire dataset exceeds 5GB and contains 15,000+ images, all structured into a comprehensive dataset.
Due to… See the full description on the dataset page: https://huggingface.co/datasets/Goku-OpenLab/gpt-image-2-prompts-datasets.GPT-Image-Edit-1.5M
GPT-Image-Edit-1.5M A Million-Scale, GPT-Generated Image Dataset
📃Arxiv | 🌐 Project Page | 💻Github
GPT-Image-Edit-1.5M is a comprehensive image editing dataset that is built upon HQ-Edit, UltraEdit, OmniEdit and Complex-Edit, with all output images regenerated with GPT-Image-1.
📣 News
[2025.08.20] 🚀 We provide a script for multi-process downloading. See Multi-process Download.
[2025.07.27] 🤗 We release GPT-Image-Edit, a state-of-the-art image editing model with… See the full description on the dataset page: https://huggingface.co/datasets/UCSC-VLAA/GPT-Image-Edit-1.5M.gpt-image-2
GPT-Image-2 Twitter Dataset
10,217 confirmed GPT-image-2.0 generated images collected from Twitter/XCollection window: April 21 – April 28, 2026 (first week post-launch)Paper: GPT-Image-2 in the Wild: A Twitter Dataset of Self-Reported AI-Generated Images from the First Week of Deployment
Overview
This dataset contains 10,217 images confirmed to be GPT-image-2.0 outputs, sourced from public Twitter/X posts in the immediate aftermath of the model's April 21, 2026 release… See the full description on the dataset page: https://huggingface.co/datasets/Scam-AI/gpt-image-2.GPT-Image-Edit-1M
GPT-Image-Edit-1M Review Artifact
GPT-Image-Edit-1M is a non-commercial research artifact for instruction-guided image editing. It contains GPT-Image-1 regenerated image-editing triplets, auditable quality-control metadata, and a 200-case human-audit package used to calibrate automated judges in the paper.
License: CC BY-NC-SA 4.0, subject to upstream dataset licenses and applicable third-party service terms.
Reviewer note. The Hugging Face Dataset Viewer shows a 400-row inspection… See the full description on the dataset page: https://huggingface.co/datasets/meimeirun/GPT-Image-Edit-1M.GPT_Watercolor_Anime_Style_Images
GPT Watercolor Anime Style Images
Dataset Description
This is a synthetic GPT-generated Watercolor Anime Style image dataset. It contains 120 image-caption pairs with transparent watercolor washes, soft ink linework, pale paper texture, muted colors, and traditional anime illustration scenes.
The images focus on soft watercolor washes, expressive linework, gentle lighting, quiet interiors, nature scenes, village streets, character studies, and calm storybook… See the full description on the dataset page: https://huggingface.co/datasets/neonforestmist/GPT_Watercolor_Anime_Style_Images.GPT_Monet_Style_Images
GPT Monet Style Images
Dataset Description
This is a synthetic GPT-generated Monet-style image dataset. It contains 100 image-caption pairs with impressionist lighting, soft broken color, painterly atmosphere, and garden, water, street, interior, and still-life compositions.
The images focus on luminous outdoor light, loose brushwork, atmospheric color, reflective water, flowers, fields, cozy scenes, cafes, village streets, and calm impressionist subjects.… See the full description on the dataset page: https://huggingface.co/datasets/neonforestmist/GPT_Monet_Style_Images.GPT_Storybook_Anime_Style_Images
GPT Storybook Anime Style Images
Dataset Description
This is a synthetic GPT-generated Storybook Anime Style image dataset. It contains 100 image-caption pairs featuring original anime-inspired characters and scenes with a warm, illustrated storybook feeling.
The images focus on expressive character moments, gentle lighting, quiet interiors, nature scenes, village streets, cozy everyday settings, and calm storybook moods. Captions commonly describe soft linework… See the full description on the dataset page: https://huggingface.co/datasets/neonforestmist/GPT_Storybook_Anime_Style_Images.gpt-image-edit-1-5m-hqeditGPT_Pointillism_Style_Images
GPT Pointillism Style Images
Dataset description
This is a synthetic GPT-generated pointillism style image dataset. It contains 70 image-caption pairs with colorful dotted texture, painterly lighting, and pointillism-inspired compositions.
The images focus on colorful stippled brushwork, luminous dotted texture, painterly lighting, cozy subjects, gardens, landscapes, interiors, and decorative still-life scenes.
Contents
images/
metadata.jsonl… See the full description on the dataset page: https://huggingface.co/datasets/neonforestmist/GPT_Pointillism_Style_Images.Output_dataset_gptImage1GPT-Image
GPT-Image Fundus Benchmark
This repository contains a processed fundus-image benchmark, training/evaluation code, experiment outputs, and model checkpoints for four ophthalmic classification directions:
Myopia / Myopic Maculopathy: MMAC, 5 classes
AMD: ADAM, binary classification
Glaucoma: AIROGS and PAPILA, binary classification
Diabetic Retinopathy (DR): IDRiD, APTOS-2019, DeepDRiD, 5 classes
The repository was prepared for experiments comparing RetFound, ResNet-50, and… See the full description on the dataset page: https://huggingface.co/datasets/MaybeRichard/GPT-Image.GPT_Photoreal_3D_Anime_Style_Images
GPT Photoreal 3D Anime Style Images
Dataset Description
This is a synthetic GPT-generated Photoreal 3D Anime Style image dataset. It contains 150 image-caption pairs with square, 1024 × 1024 PNG images.
The images cover individual and group portraits, everyday activities, animals, objects, environments, and fantasy scenes. Captions describe anime-inspired forms, photorealistic surface textures, and physically based 3D lighting. Every caption begins with Photoreal… See the full description on the dataset page: https://huggingface.co/datasets/neonforestmist/GPT_Photoreal_3D_Anime_Style_Images.gpt-image-mega-4k
4000 Highly curated GPT-IMAGE 1 Images captioned with QWEN-3-VL
This is a high-quality aesthetically pleasing image dataset containing 4,000 carefully curated & captioned GPT-IMAGE 1 images collected for training and fine-tuning text-to-image models.
Every image was selected for visual clarity, composition, lighting, and stylistic appeal, making this dataset one of the best out there.
Captioned with: Qwen3-VL
Resolution: 1024 X 1536
Image format: PNG
Total image: 4000
Use case:… See the full description on the dataset page: https://huggingface.co/datasets/a3xrfgb/gpt-image-mega-4k.liars-dividend-GPT-image-2-photorealgptImagesRandomTopicsgpt-image-2-5-prompt-atlas
GPT Image 2.5 Prompt Atlas
50 prompts for GPT Image 2.5, each with the frame it produced, as a single machine-readable file.
Every record pairs the exact prompt text with the image files that came back, so a prompt can be read, re-run and compared without reconstruction.
Files
File
Contents
prompts.jsonl
50 records, one per prompt
Each record:
{
"id": "showcase/13-silhouette-universe-poster",
"group": "showcase",
"slug":… See the full description on the dataset page: https://huggingface.co/datasets/callirra/gpt-image-2-5-prompt-atlas.gpt-image-edit-omniedit-ultraeditgpt-image-edit-omniedit-ultraedit-part2GPT-4o-Design-Images-QwenVL_2_5_Captionedimage_style_transfer_GPTImage2
Multi-Style Stylized Image Pairs
A paired-image dataset where every source photograph is re-rendered into four
distinct artistic styles by the same model, producing a one-to-many style
transfer benchmark with consistent composition across styles.
Styles
For each raw image, the dataset provides four stylized variants generated
from the same source:
Column
Style description
raw
The original source photograph.
American_Cartoon
Modern American cartoon… See the full description on the dataset page: https://huggingface.co/datasets/yufan/image_style_transfer_GPTImage2.gpt-image-edit-benchmark-results
GPT-Image-Edit — Benchmark Results
This repository contains evaluation results of GPT-Image-Edit across four standard image-editing benchmarks. All scores were computed using the official evaluation scripts provided by each benchmark.
📊 Benchmarks
Benchmark
Metrics
Folder
GEdit-EN
12 editing categories + Avg
gedit/
Complex-Edit
IF, IP, PQ, Overall
complex_edit/
ImgEdit-Full
10 editing operations + Overall
imgedit/
OmniContext
Contextual edit scores… See the full description on the dataset page: https://huggingface.co/datasets/UCSC-VLAA/gpt-image-edit-benchmark-results.Gpt_Style_Ghibli_Images
📦 Bulk Download 7z — GPT4 Style Ghibli Images Dataset
🔗 https://huggingface.co/datasets/9r4n4y/Gpt_Style_Ghibli_Images/blob/main/GPT4_Style_Ghibli_Images_full_images.7z
AI-Generated Ghibli Images Dataset
This repository contains a collection of AI-generated images in the style of Studio Ghibli. These images showcase various subjects, characters, and landscapes rendered in the distinctive artistic style associated with Studio Ghibli animation films.
Dataset… See the full description on the dataset page: https://huggingface.co/datasets/9r4n4y/Gpt_Style_Ghibli_Images.VIBE-GPT-Imageaas_benchmark_gpt-gpt-image-1.5GPT-Image-PICABenchgpt-images-2-5-prompts
GPT Images 2.5 - Comprehensive Prompt & Staging Benchmark
Welcome to the official prompt staging and benchmarking repository for GPT images 2.5.
Overview
GPT images 2.5 is a next-generation AI image generation and conversational editing platform powered by the Flare engine. It supports conversational inpainting, intelligent aspect-ratio re-framing, and crystal-clear 4K output.
Official Web Application: https://images25.art
Engine: Flare V2.5
Features: Free… See the full description on the dataset page: https://huggingface.co/datasets/ryanellis42675/gpt-images-2-5-prompts.aas_benchmark-gpt-image-mini-highgpt-image-2-zq10k-2048
GPT-Image-2 ZQ10K 2048 — Photometrically Corrected v2
This version replaces the earlier low-key release with the 10,000-image dataset
used by the latest APT experiment. Images were generated at 2048 x 2048 by
openai/openai/gpt-image-2 from the semantic-lighting-balanced v2 prompts, then corrected
with M4 quantile PCHIP, luminance-only, shrink 0.8.
Mean luminance changed from 0.41873 to
0.50619; the paired full target mean was
0.53295. Original corrected PNG bytes are embedded… See the full description on the dataset page: https://huggingface.co/datasets/Kontext-Style/gpt-image-2-zq10k-2048.gpt-generated-image-collection
GPT-Generated Image Collection
A growing collection of contributor-provided GPT-generated PNG images. The first release contains 216 original PNG files. Exact generation-model versions have not been independently verified; the generation_model field is therefore null rather than an inferred model name.
For the Chinese version of this card, see README.zh-CN.md.
Layout
README.md English dataset card
README.zh-CN.md Chinese… See the full description on the dataset page: https://huggingface.co/datasets/winrisef/gpt-generated-image-collection.gpt-image
