
A cost-efficient version of GPT Image 1
Frequently asked questions
What is gpt-image-1-mini?
gpt-image-1-mini is a cost-efficient image generation model developed by OpenAI. It is a natively multimodal model that accepts both text prompts and image inputs to produce high-quality images. It is available on Craisee through the Replicate provider and is designed for creators and developers who need scalable image generation at a lower cost.
How does gpt-image-1-mini differ from GPT Image 1?
gpt-image-1-mini is a smaller, more affordable version of OpenAI's full GPT Image 1 model. It offers the same core capabilities — text-to-image and image-to-image generation — but at a reduced cost, making it ideal for teams and developers who need high-volume image generation without the premium pricing of the full-sized model.
What image formats does gpt-image-1-mini support?
gpt-image-1-mini supports PNG, JPEG, and WebP output formats, covering a wide range of web and print delivery needs. It also supports transparent, opaque, or automatically determined backgrounds, making it suitable for product design, UI/UX assets, and other creative workflows requiring flexible image formatting.
Can gpt-image-1-mini generate multiple images at once?
Yes, gpt-image-1-mini supports batch image generation, producing up to 10 images in a single run. This enables rapid iteration, variation exploration, and efficient creative workflows for developers and designers who need multiple image options quickly without running separate generation requests.
What aspect ratios does gpt-image-1-mini support?
gpt-image-1-mini supports 1:1 (square), 3:2 (landscape), and 2:3 (portrait) aspect ratios. These cover common creative formats including social media posts, portrait photography, and landscape layouts, giving creators flexibility to match their output to specific platform or design requirements.
How do I use gpt-image-1-mini for image-to-image generation?
To use gpt-image-1-mini for image-to-image generation, provide one or more input images alongside a text prompt. The model processes both visual and textual context simultaneously, enabling style transfer, guided editing, and reference-based generation. Input fidelity control lets you specify how closely the output should match the style or facial features of reference images.
What quality settings are available in gpt-image-1-mini?
gpt-image-1-mini offers four quality tiers: low, medium, high, and auto. These settings allow users to balance output fidelity against generation cost and speed. The auto setting lets the model determine the optimal quality level, while manual settings give developers precise control over resource usage and image quality.
Where can I access gpt-image-1-mini?
gpt-image-1-mini is available on Craisee through the Replicate provider. It is accessible under the model slug openai/gpt-image-1-mini and supports both text-to-image and image-to-image workflows. Craisee provides a platform for creators, developers, and teams to use this model at scale with cost-efficient pricing.



