All Models
GPT Image 2
GPT Image 2 generates and edits stills from natural-language prompts, with quality and resolution controls.
Call the API endpoint to get the latest, detailed pricing and pricing estimates.
Call the API endpoint to get the latest, detailed pricing and pricing estimates.
Call the API endpoint to get the latest, detailed pricing and pricing estimates.
GPT Image 2
GPT Image 2 is OpenAI’s stills model. It generates and edits from natural-language prompts. Quality is low, medium, or high, default medium. Resolution is 1K, 2K, or 4K, default 1K. Aspect ratios cover square through 21:9 and 9:21, including 1:2, 2:1, 1:3, and 3:1. Output is PNG, JPEG, or WebP. Edits take a list of images.
It follows long written briefs: who is in the scene, what they are wearing, what is on the table, what should be legible in the corner. Quality low is for finding the composition. Raise quality and resolution once the prompt is stable. There is no decompose, vectorize, or upscale operation.
What it's good for
General stills from a written description: editorial art, presentation images, marketing frames, concept work. It is not a motion tool and not a vector tool. Instruction following is the point.
Model ID
openai/gpt-image-2
Output Type
Image
Pricing
From $0.01
Provider
OpenAI
Hosting
Protoface
GPT Image 2
GPT Image 2 is OpenAI’s stills model. It generates and edits from natural-language prompts. Quality is low, medium, or high, default medium. Resolution is 1K, 2K, or 4K, default 1K. Aspect ratios cover square through 21:9 and 9:21, including 1:2, 2:1, 1:3, and 3:1. Output is PNG, JPEG, or WebP. Edits take a list of images.
It follows long written briefs: who is in the scene, what they are wearing, what is on the table, what should be legible in the corner. Quality low is for finding the composition. Raise quality and resolution once the prompt is stable. There is no decompose, vectorize, or upscale operation.
What it's good for
General stills from a written description: editorial art, presentation images, marketing frames, concept work. It is not a motion tool and not a vector tool. Instruction following is the point.