GPT-4o Image API

Leverage OpenAI’s omnimodal visual engine. The GPT-4o Image API delivers pixel-perfect text, cinematic realism, and native inpainting capabilities—optimized by EvoLink for cost and speed.

Configure Request

Response

Click "Generate" to see results

GPT-4o Image API: Precision visuals, readable text

Generate and edit assets with the industry's most accurate text rendering model. From 1024px marketing visuals to complex outpainting, streamline your workflow with native C2PA provenance.

What can you build with the GPT-4o Image API?

Typographically accurate designs

Finally, an AI that spells correctly. Generate posters, UI mocks, and logos with reliable Latin text rendering.

View examples

Commercial-grade photorealism

Create studio-quality product shots and portraits with advanced lighting and texture control.

See quality

Native Inpainting & Outpainting

Edit existing images programmatically. Mask specific regions to modify objects or expand canvas boundaries seamlessly.

Test editing

Why developers choose the GPT-4o Image API

Surpassing DALL·E 3 in text fidelity and instruction following, the GPT-4o Image API offers a unified endpoint for generation and editing at a competitive price point (~$0.035 via EvoLink).

Superior Instruction Following

The model understands complex, multi-step prompts better than any previous diffusion model, reducing retry costs.

Cost-Efficient Scaling

Access premium visual capabilities at ~$0.035 per 1024px render through EvoLink's optimized routing infrastructure.

Trust & Safety Standard

Every generation includes C2PA metadata for provenance and adheres to enterprise-grade safety filters.

How to integrate the GPT-4o Image API

From prompt to pixel-perfect result in three simple requests.

Step 1 — Configure Payload

Set your model to `gpt-4o-image`, define dimensions (default 1024x1024), and choose 'generate' or 'edit' endpoints.

Step 2 — Send Contextual Prompts

Submit detailed text instructions. For edits, include your base image and a transparency mask for the target area.

Step 3 — Retrieve & Deploy

Receive high-fidelity URLs or Base64 data. Process C2PA metadata and integrate the asset directly into your app.

Open API Docs

Core API Capabilities

Precision

Text Rendering Engine

Drastically reduced spelling errors compared to DALL·E 3.

Editing

Advanced Inpainting

Programmatic masking to add, remove, or modify image regions.

Creativity

Style Versatility

Switch between photorealism, 3D render, and line art via prompts.

Quality

High-Res Output

Native 1024px generation with scalable quality for print.

Compliance

Enterprise Safety

Built-in content moderation and provenance watermarking.

Value

Optimized Pricing

Route via EvoLink to access rates around $0.035 per generation.

Compare Image Generation Models

Choose the right tool for your specific visual needs

Model	Duration	Resolution	Price	Strength
GPT-4o Image API	N/A	1024px (Native)	~$0.035/image	Superior text rendering, native inpainting, highest prompt fidelity.
Nano Banana 2	N/A	1024px Base	~$0.12/image	Strong multilingual support; fast throughput for bulk tasks.
Wan 2.5 T2I	N/A	Variable (Token)	~$0.03/image	Flexible aspect ratios & experimental style batching.

GPT-4o Image API - Frequently Asked Questions

Everything you need to know about the product and billing.

Through EvoLink routing, the effective cost is approximately $0.035 per 1024px image, making it a cost-effective alternative to standard DALL·E 3 list pricing.

GPT-4o Image API uses an omnimodal architecture that significantly improves text rendering accuracy, prompt adherence, and enables native inpainting/outpainting capabilities that DALL·E 3 lacks.

Yes. The endpoint accepts a base image and a mask image, allowing you to use text prompts to edit or replace specific areas of the visual seamlessly.

The standard native generation is 1024x1024 pixels. For higher requirements, EvoLink offers upscaling pipelines to reach 2048px or 4K.

Yes, you own the images you generate. Additionally, the API includes C2PA metadata to verify provenance and safety filters to prevent policy violations.

While highly detailed, the GPT-4o Image API is optimized for speed. Exact latency depends on server load, but EvoLink routing prioritizes the fastest available path.

Yes, the API supports various aspect ratios including square (1:1), wide (16:9), and tall (9:16) to suit different platform needs.

API Reference

Select endpoint

Authentication

All APIs require Bearer Token authentication.

Header

Authorization: 
Bearer YOUR_API_KEY

Get API Key

POST

/v1/images/generations

Generate Image

Create an image generation task using text prompts or reference images. Supports text-to-image, image-to-image, and image editing modes.

Asynchronous processing mode, use the returned task ID to .

Generated image links are valid for 24 hours, please save them promptly.

Request Parameters

modelstringRequiredDefault: gpt-4o-image

Image generation model name.

Examplegpt-4o-image

promptstringRequired

Prompt describing the image to be generated or how to edit the input image.

Notes

Limited to 2000 tokens

ExampleA cat playing in the grass

sizestringOptional

Size of the generated image. Supports ratio format and pixel format.

Value	Description
1:1	Square ratio
2:3	Portrait ratio
3:2	Landscape ratio
1024x1024	Square (1024×1024)
1024x1536	Portrait (1024×1536)
1536x1024	Landscape (1536×1024)

Example1024x1024

nintegerOptional

Number of images to generate.

Value	Description
1	Generate 1 image
2	Generate 2 images
4	Generate 4 images

Notes

Pre-charged based on n value, actual charge based on generated count

Example1

image_urlsarrayOptional

Reference image URL list for image-to-image and image editing functions.

Notes

Max 5 images per request
Max size: 10MB per image
Formats: .jpeg, .jpg, .png, .webp
URLs must be directly accessible by the server

Examplehttps://example.com/image1.png

mask_urlstringOptional

Mask image URL for image editing.

Notes

Mask and reference image must be PNG format with same dimensions
Max size: 4MB
Cannot be used when image_urls contains more than 1 image

Examplehttps://example.com/mask.png

callback_urlstringOptional

HTTPS callback address after task completion.

Notes

Triggered on completion, failure, or cancellation
Sent after billing confirmation
HTTPS only, no internal IPs
Max length: 2048 chars
Timeout: 10s, Max 3 retries

Examplehttps://your-domain.com/webhooks/image-task-completed

Request Example

{
  "model": "gpt-4o-image",
  "prompt": "A beautiful sunset over the ocean with vibrant colors",
  "size": "1024x1024",
  "n": 1
}

Response Example

{
  "created": 1757165031,
  "id": "task-unified-1757165031-uyujaw3d",
  "model": "gpt-4o-image",
  "object": "image.generation.task",
  "progress": 0,
  "status": "pending",
  "task_info": {
    "can_cancel": true,
    "estimated_time": 45
  },
  "type": "image",
  "usage": {
    "billing_rule": "per_call",
    "credits_reserved": 2.5,
    "user_group": "default"
  }
}