GPT-Image-2 API — OpenAI's image model at half the price, OpenAI-compatible

GPT-Image-2 is OpenAI's latest native image model — strong on prompt understanding and text rendering. Kunavo serves it through the OpenAI-compatible images endpoint, at about half of OpenAI's list price, on one bearer token. If your code can call the OpenAI images API, it can call GPT-Image-2 by changing only the base_url. It's the same interface devs search for as the “GPT image API”.

Call the GPT-Image-2 API

Generate a key at /app/keys, keep the OpenAI SDK, and point base_url at Kunavo. Text-to-image is POST /v1/images/generations:

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["KUNAVO_API_KEY"],
    base_url="https://api.kunavo.com/v1",
)

img = client.images.generate(
    model="gpt-image-2",
    prompt="a product hero shot of a matte-black espresso machine, studio light",
    size="1024x1024",
)
print(img.data[0].url)  # temporary URL (~24h) — download and re-host

Returned URLs are temporary (~24h) — download and re-host on your own CDN. Full parameters (size, resolution, quality, async + webhooks) are in the images API reference.

Pricing

On Kunavo, gpt-image-2 is about half of OpenAI's list, and the price is flat across resolution tiers. Live rates are on the model page.

Edit an image (image-to-image)

POST /v1/images/edits takes a prompt plus a source image — an https URL or a data: base64 URI — and returns the edited result. Use gpt-image-2-edit for GPT-Image-2 image-to-image:

# Image-to-image / edit: pass a source image (https URL or data: base64 URI)
curl https://api.kunavo.com/v1/images/edits \
  -H "Authorization: Bearer $KUNAVO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-image-2-edit",
    "prompt": "add steam rising from the cup, keep the machine",
    "image": "https://example.com/source.png"
  }'

GPT-Image-2 vs Nano Banana

Reach for GPT-Image-2 when prompt fidelity and text inside the image matter most; reach for Nano Banana when you want the cheapest high-volume generation. Both run on the same endpoint and key, so you can A/B them by changing one string. For the full multimodal picture — image → video → audio pipelines — see the multimodal AI guide.

FAQ

What is the GPT-Image-2 API?

OpenAI's latest image model, served on Kunavo's OpenAI-compatible /v1/images/generations endpoint at about half of list price.

How much does it cost?

About $0.0633 per image versus OpenAI's ~$0.1266 — roughly 50% off, flat across 1K/2K/4K, billed per image.

How do I get a GPT-Image-2 API key?

Sign up, create a key at /app/keys, and set base_url=https://api.kunavo.com/v1 — no separate OpenAI account. See the quickstart.

Can I edit images, not just generate them?

Yes — POST /v1/images/edits with a source image does image-to-image. Details in the images reference.