GENERATION API CALLS

Image Model Generation

OpenAI

This page is auto-generated from model configurations. Last updated: 2026-10-01.

This reference lists all available OpenAI image generation models and their parameters. Use these parameter names when calling the Generation API.

GPT Image
GPT Image 1.5
GPT Image 2
GPT Image 2.5 Flare
GPT Image 2.5 Sunburst

GPT Image

OpenAI’s GPT Image 1 model for image editing and generation

Model ID: model_openai-gpt-image-1-editing

Capabilities: txt2img, img2img

LLM Markdown: https://app.scenario.com/api/models/model_openai-gpt-image-1-editing/markdown

Parameter	Type	Required	Default	Min	Max	Allowed Values	Description
`referenceImages`	file_array	No	-	-	-	-	Reference images for style or content guidance
`prompt`	string	Yes	-	-	-	-	Text prompt for image editing
`numOutputs`	number	No	`1`	1	10	-	Number of images to generate
`aspectRatio`	string	No	`auto`	-	-	`auto`, `1:1`, `3:2`, `2:3`	Aspect ratio for the generated image
`inputFidelity`	string	No	`high`	-	-	`low`, `high`	Preserve details from input images
`quality`	string	No	`high`	-	-	`high`, `medium`, `low`	Generation quality
`background`	string	No	`auto`	-	-	`transparent`, `opaque`, `auto`	Background transparency

GPT Image 1.5

OpenAI’s GPT Image 1.5 model for image editing

Model ID: model_openai-gpt-image-1-5-editing

Capabilities: txt2img, img2img

LLM Markdown: https://app.scenario.com/api/models/model_openai-gpt-image-1-5-editing/markdown

Parameter	Type	Required	Default	Min	Max	Allowed Values	Description
`referenceImages`	file_array	No	-	-	-	-	Reference images for style or content guidance
`prompt`	string	Yes	-	-	-	-	Text prompt for image editing
`numOutputs`	number	No	`1`	1	10	-	Number of images to generate
`aspectRatio`	string	No	`auto`	-	-	`auto`, `1:1`, `3:2`, `2:3`	Aspect ratio for the generated image
`inputFidelity`	string	No	`high`	-	-	`low`, `high`	Preserve details from input images
`quality`	string	No	`high`	-	-	`high`, `medium`, `low`	Generation quality
`background`	string	No	`auto`	-	-	`transparent`, `opaque`, `auto`	Background transparency

GPT Image 2

OpenAI GPT Image 2 for high-fidelity image generation and editing (Modal)

Model ID: model_openai-gpt-image-2

Capabilities: txt2img, img2img

LLM Markdown: https://app.scenario.com/api/models/model_openai-gpt-image-2/markdown

Parameter	Type	Required	Default	Min	Max	Allowed Values	Description
`prompt`	string	Yes	-	-	-	-	Text description of desired generation or edit.
`referenceImages`	file_array	No	-	-	-	-	Reference images for editing. Each image must be under 50 MB.
`mask`	file	No	-	-	-	-	PNG mask for inpainting, same dimensions as the first reference image. OpenAI uses the alpha channel only: fully transparent pixels are edited, fully opaque pixels are preserved. RGB white/black is ignored if there is no alpha. Export PNG with transparency — do not flatten. If you paint a white-on-black mask, convert the white (edit) region to transparent.
`numOutputs`	number	No	`1`	1	10	-	Number of images to generate
`width`	number	No	-	16	3840	-	Output width in pixels.
`height`	number	No	-	16	3840	-	Output height in pixels.
`quality`	string	No	`auto`	-	-	`auto`, `high`, `medium`, `low`	Generation quality
`background`	string	No	`auto`	-	-	`auto`, `opaque`	Background option. Transparent backgrounds are not supported on GPT Image 2; use GPT Image 2.5 Flare or GPT Image 2.5 Sunburst.

GPT Image 2.5 Flare

OpenAI GPT Image 2.5 Flare for fast, high-quality everyday image generation and editing.

Model ID: model_openai-gpt-image-2-5-flare

Capabilities: txt2img, img2img

LLM Markdown: https://app.scenario.com/api/models/model_openai-gpt-image-2-5-flare/markdown

Parameter	Type	Required	Default	Min	Max	Allowed Values	Description
`prompt`	string	Yes	-	-	-	-	Describe the image you want to create, or how to edit your reference images. Plain, natural language works well.
`referenceImages`	file_array	No	-	-	-	-	Images to edit or build from (up to 10). Required if you provide a mask.
`mask`	file	No	-	-	-	-	An optional PNG that marks which parts of the first reference image to change. It uses transparency: fully transparent areas are edited, fully opaque areas are kept. Export a PNG with real transparency (don’t flatten it). If you’ve drawn a white-on-black mask, convert the white “edit” areas to transparent first.
`numOutputs`	number	No	`1`	1	10	-	How many images to create from the same prompt (1-10).
`width`	number	No	-	16	3840	-	The output width in pixels.
`height`	number	No	-	16	3840	-	The output height in pixels.
`quality`	string	No	`auto`	-	-	`auto`, `low`, `medium`, `high`, `xhigh`, `max`	How much effort goes into the image. Higher levels look better but cost more; Auto lets the model choose.
`background`	string	No	`auto`	-	-	`auto`, `opaque`, `transparent`	Whether the background is solid (Opaque) or see-through (Transparent). Auto lets the model decide.

GPT Image 2.5 Sunburst

OpenAI GPT Image 2.5 Sunburst, the most capable GPT Image model for generation and editing where precision matters most.

Model ID: model_openai-gpt-image-2-5-sunburst

Capabilities: txt2img, img2img

LLM Markdown: https://app.scenario.com/api/models/model_openai-gpt-image-2-5-sunburst/markdown

Parameter	Type	Required	Default	Min	Max	Allowed Values	Description
`prompt`	string	Yes	-	-	-	-	Describe the image you want to create, or how to edit your reference images. Plain, natural language works well.
`referenceImages`	file_array	No	-	-	-	-	Images to edit or build from (up to 10). Required if you provide a mask.
`mask`	file	No	-	-	-	-	An optional PNG that marks which parts of the first reference image to change. It uses transparency: fully transparent areas are edited, fully opaque areas are kept. Export a PNG with real transparency (don’t flatten it). If you’ve drawn a white-on-black mask, convert the white “edit” areas to transparent first.
`numOutputs`	number	No	`1`	1	10	-	How many images to create from the same prompt (1-10).
`width`	number	No	-	16	3840	-	The output width in pixels.
`height`	number	No	-	16	3840	-	The output height in pixels.
`quality`	string	No	`auto`	-	-	`auto`, `low`, `medium`, `high`, `xhigh`, `max`	How much effort goes into the image. Higher levels look better but cost more; Auto lets the model choose.
`background`	string	No	`auto`	-	-	`auto`, `opaque`, `transparent`	Whether the background is solid (Opaque) or see-through (Transparent). Auto lets the model decide.