GPT-Image-1

Image GenerationReference EditingMultimodal Image

GPT Image 1 is OpenAI’s previous-generation, natively multimodal image model for text-to-image creation and reference-guided editing. APIXO continues to expose a dedicated route with prompt input, up to five reference images, square or portrait/landscape framing, and one, two, or four outputs per task.

Inputs

Text + Up to 5 Images

APIXO Price

$0.35/ Image

APIXO Ratios

1:1 / 2:3 / 3:2

Output Count

1 / 2 / 4 Images

Released

Apr 23, 2025

Loading workspace...

Capabilities

Create Images from Language and Visual Context

Prompt-Based Generation

Generate an image from natural-language direction covering subject, composition, style, lighting, camera treatment, and visible details. GPT Image 1 was designed to follow detailed creative instructions instead of depending on terse keyword-style prompts.

Multi-Image Guidance

APIXO accepts up to five public reference-image URLs in one task. Use them to provide visual context for subjects, products, materials, composition, or style while describing the intended relationship between the inputs in the prompt.

Instruction-Based Editing

Supply an existing image and describe the desired transformation in ordinary language. GPT Image 1 can alter visual elements, restyle material, or construct a new composition informed by the source without requiring a separate model family.

Typography Generation

GPT Image 1 improved the ability to place requested words inside generated visuals compared with earlier DALL·E workflows. It supports concepts for posters, labels, diagrams, and social graphics, though all rendered text still requires human review.

Broad Style Range

Create photography-like scenes, illustrations, graphic compositions, diagrams, and other visual treatments through one model. Reference images and detailed prompts can help align the output with an established art direction or campaign concept.

Batch Creative Variants

APIXO can return one, two, or four images from a task. Multiple outputs help teams compare interpretations of the same prompt or reference set before selecting a direction for refinement or downstream production.

API Details

GPT Image 1 Options on APIXO

Verified input, framing, quantity, and delivery limits for APIXO’s dedicated GPT Image 1 route.

1–10,000 Characters

Prompt Length

0–5 Images

Reference Count

JPG / PNG / WebP

Input Formats

Up to 10 MB Each

Reference Size

1 / 2 / 4 Images

Images per Task

Polling or Webhook

Delivery

Production Uses

Apply GPT Image 1 to Existing Creative Pipelines

Campaign Design

Marketing Asset Concepts

Generate hero-image directions, social graphics, editorial illustrations, and campaign concepts from detailed creative briefs. Request several variants when stakeholders need multiple interpretations before selecting an asset for refinement in conventional design software.

Product Content

Reference-Guided Product Scenes

Combine product photographs with contextual or stylistic references to explore new settings and presentation ideas. This supports e-commerce concepts, merchandising tests, and advertising drafts, while final outputs should be checked for accurate product geometry, branding, and labels.

Visual Communication

Posters and Diagram Drafts

Create early layouts containing headlines, labels, symbols, and explanatory imagery. GPT Image 1’s typography capabilities make it useful for visual ideation, but generated words, figures, and factual content should be verified and typeset separately for final publication.

Creative Exploration

Multi-Reference Style Studies

Supply several visual references to explore how subjects, materials, palettes, and composition might work together. Designers can compare one, two, or four generated directions before developing the strongest concept into a production-ready asset.

Notes & FAQ

GPT Image 1 Availability and Constraints

Operational Notes

GPT Image 1 predates GPT Image 2 but remains available through APIXO’s dedicated route.

APIXO exposes one GPT Image 1 routing mode; it does not group GPT Image 1.5, GPT Image 2, or DALL·E models.

APIXO documents reference-guided creation but does not expose OpenAI mask input or explicit inpainting controls on this route.

APIXO does not publish controls for quality, transparency, output format, compression, moderation level, or input fidelity for this endpoint.

Reference inputs must be public, directly reachable JPG, PNG, or WebP files no larger than 10 MB each.

Generated text, faces, fine details, and reference identity may require additional attempts or manual correction before production use.

Frequently Asked Questions

Yes. APIXO publicly lists a dedicated GPT Image 1 endpoint and Playground workflow. GPT Image 2 is a separate model and endpoint, so existing GPT Image 1 integrations can continue to use the configuration documented on this page.

APIXO supports reference-guided generation using up to five image URLs and a required prompt. This enables instruction-based transformations and visual guidance, but APIXO does not document the dedicated mask parameter available through OpenAI’s native image-editing interface.

Transparency is an upstream GPT Image 1 capability, but APIXO’s public schema for this endpoint does not expose a background or output-format parameter. Do not depend on transparent PNG output through APIXO unless the live schema explicitly adds that option.

APIXO does not publish OpenAI’s low, medium, high, or auto quality controls for this route. It exposes three aspect-ratio choices—1:1, 2:3, and 3:2—rather than the upstream API’s explicit pixel-size parameter.

APIXO charges $0.35 for each generated image. A task requesting two or four outputs is billed for the corresponding number of images. This APIXO unit price is separate from OpenAI’s native token- and quality-based pricing structure.