GPT Image 2 API Guide: Features, Pricing & Code

OpenAI released GPT Image 2 on April 21, 2026, as the model behind ChatGPT Images 2.0, replacing both DALL-E 3 and the shorter-lived GPT Image 1.5. It's the first OpenAI image model with a built-in reasoning step, and that shows up in cleaner text rendering, better multi-image consistency, and more accurate instruction following. It's live on Apiframe now: one key, one endpoint, the same billing and concurrency you already use for every other model on the platform.

This guide covers what's actually new, one real limitation worth knowing before you build, then walks through generating with the model, including pricing you can plan a budget around.

What's New in GPT Image 2

Reasoning built in. Before generating an image, the model can plan the layout, search the web for reference material, and check its own output. No earlier OpenAI image model had this "thinking" step.

Up to 8 images from one prompt. It produces a matching set of images in a single request, keeping characters and objects consistent across all of them. Useful for storyboards and brand campaigns that used to need separate, unrelated generations stitched together.

State-of-the-art text rendering. Dense and multilingual text renders cleanly across several scripts, reliable enough for posters, packaging, menus, and UI labels, areas where earlier models tended to garble small or non-Latin text.

Precise instruction following. OpenAI describes it as a major step up in following detailed prompts and placing objects accurately within a scene.

Higher resolution and flexible editing. Up to 2K resolution through the API, plus multi-turn editing that keeps identity and context consistent across changes.

One real limitation worth knowing upfront

Apiframe's own marketing copy for GPT Image 2 says "up to 2K resolution via the API," but the actual parameters don't include a resolution or size selector, only three aspect ratios (1:1, 3:2, 2:3), with resolution set indirectly by the quality setting rather than something you can request directly. If your project needs explicit control over output resolution, or aspect ratios beyond those three, that control isn't currently available on Apiframe's gpt-image-2 endpoint. Worth knowing before you design around a specific resolution target.

Getting Started on Apiframe

GPT Image 2 runs through the same API key and billing as every other model on Apiframe, including Flux 2 Pro, Nano Banana Pro, and Kling. It's not a separate integration, just one more model on the integration you may already have. Requests are asynchronous: submit a job, then poll for the result or receive it through a webhook. No OpenAI account required.

If you're starting from scratch, sign up for free credits, grab an API key, and you can be generating within minutes. The full parameter reference lives in the GPT Image 2 docs, or you can try the model directly in Apiframe Studio without writing any code. For a broader look at how image APIs work in general, see our AI Image API guide for developers.

Full Code Walkthrough

1. Text-to-image

curl -X POST https://api.apiframe.ai/v2/images/generate \
  -H "X-API-Key: afk_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-image-2",
    "prompt": "Infographic explaining how photosynthesis works, bold sans-serif labels, light academic color palette",
    "gptImage2Params": {
      "quality": "high",
      "aspect_ratio": "3:2"
    }
  }'

The same call in Python:

import requests

response = requests.post(
    "https://api.apiframe.ai/v2/images/generate",
    headers={
        "X-API-Key": "afk_your_api_key_here",
        "Content-Type": "application/json",
    },
    json={
        "model": "gpt-image-2",
        "prompt": "Infographic explaining how photosynthesis works, bold sans-serif labels, light academic color palette",
        "gptImage2Params": {
            "quality": "high",
            "aspect_ratio": "3:2",
        },
    },
)
print(response.json())  # { "jobId": "...", "status": "QUEUED" }

2. Image editing and combination

Pass one or more input_images to edit an existing image, or combine subjects and styles into a single output:

response = requests.post(
    "https://api.apiframe.ai/v2/images/generate",
    headers={"X-API-Key": "afk_your_api_key_here"},
    json={
        "model": "gpt-image-2",
        "prompt": "Change the red hat to a light-blue velvet beret. Preserve the subject's face, pose, and lighting.",
        "gptImage2Params": {
            "input_images": ["https://example.com/portrait.jpg"],
            "quality": "auto",
        },
    },
)

The model processes input images at high fidelity automatically. There's no separate fidelity setting to adjust, unlike GPT Image 1.5. More on that difference in GPT Image 1.5 vs 2: Which One to Use. If you're building an editing-focused workflow, our AI Image Editing API guide covers the broader landscape.

3. Poll for the result

The submission returns a 202 with a jobId. Poll GET /v2/jobs/{id} until the status is COMPLETED, or supply a webhook_url instead:

import time

while True:
    job = requests.get(
        f"https://api.apiframe.ai/v2/jobs/{job_id}",
        headers={"X-API-Key": "afk_your_api_key_here"},
    ).json()
    if job["status"] in ("COMPLETED", "FAILED"):
        break
    time.sleep(3)

print(job["result"])  # { "images": ["https://cdn2.apiframe.ai/images/..."] }

Average completion time is around 15 seconds, fast enough that polling is practical even without a webhook, though webhooks still scale better for high-volume production use.

Pricing

GPT Image 2 bills a flat credit cost per image, based on the quality setting, and the total is multiplied by number_of_images:

Credits cost a flat $0.01 each on every plan, whether you buy them as a pay-as-you-go top-up or as part of a monthly subscription. Plans differ in monthly credit volume and concurrency limits, not in the per-credit rate, so the per-image cost above is what you pay regardless of tier. See the full plan breakdown on the pricing page.

For example, a request with quality: "medium" and number_of_images: 4 costs 7 × 4 = 28 credits, regardless of which plan you're on. If you want to compare costs across image models more broadly, see AI image API pricing in 2026: what you actually pay per image.

GPT Image 2 vs GPT Image 1.5

A cheaper, still-current sibling model, gpt-image-1.5, remains available on Apiframe. It's mainly worth considering if you need transparent background output, which GPT Image 2 dropped in favor of auto or opaque backgrounds only. See the full breakdown of what changed between the two versions in GPT Image 1.5 vs 2: Which One to Use.

Use Cases

Storyboards and brand campaigns. Consistency across up to 8 images makes multi-panel sequences and campaign sets practical without stitching together separate, unrelated generations.

Posters, packaging, and UI mockups. Text rendering is strong enough to trust for the final asset rather than a placeholder you redo by hand. Pair it with Flux 2 Pro if you want to compare photorealistic styles on the same project.

Product edits. Multi-turn editing preserves identity and composition across changes, useful for refining a single asset without regenerating it from scratch each time.

Educational and informational content. Infographics and diagrams benefit directly from the reasoning step, which plans the layout before rendering a single pixel.

If you're comparing image models more broadly before committing, see 8 Best AI Image Generator APIs in 2026 or Best AI Image APIs in 2026: Midjourney, Flux, Stable Diffusion & More Compared. And if character consistency across a series of images matters to your project, our guide on keeping AI-generated characters consistent walks through the techniques in more depth.