GPT Image 2 Prompts Dataset: 15,000+ Image Generation Prompts with Metadata

A collection of over 15,000 prompts used for OpenAI's GPT Image 2 model, paired with the resulting generated images. The dataset, created by Goku-OpenLab, exceeds 5GB in size and includes full metadata and preview images. It was last updated on Hugging Face in July 2026.

Use Cases

• Training or fine-tuning text-to-image models based on the prompt-image pairs.
• Analyzing the relationship between prompt text and generated image content based on the provided metadata.
• Benchmarking image generation models using the collection of prompts and associated outputs.
• Studying prompt engineering techniques based on the large corpus of example prompts.

Strengths

• Contains over 15,000 prompt-image pairs, providing a substantial corpus for analysis.
• Dataset size exceeds 5GB, suggesting it includes high-resolution preview images.
• Includes full metadata for the prompts and images, which may aid in structured analysis.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last updated 2026-07-21 05:33:27; freshness should be verified.

Provenance

Source

Goku-OpenLab on Hugging Face.

Collection Method

Likely collected from usage of OpenAI's GPT Image 2 model, though the specific gathering method is not detailed.

Freshness

Last updated 2026-07-21 05:33:27.