Native OpenAI quality
Direct access to gpt-image-2 output quality through AllToken, with no extra quality layer in the way.
Studio is live
Image 2 — OpenAI's flagship image model, on AllToken.
Generate production-quality images through the same OpenAI-compatible API surface. Async tasks, idempotency, one AllToken API Key.
Image models
Capabilities
Production image generation behind one API surface.
Image 2 gives teams a practical route to OpenAI image quality without adding another provider workflow.
One API surface
Use the same AllToken API Key for chat, image, and video workloads instead of managing separate OpenAI billing.
Async + idempotency
Submit a task, poll the result, respect the 30-minute TTL, and use Idempotency-Key for safe retries.
Region availability
Region availability is handled by the upstream provider; AllToken keeps billing and account access unified.
API
Create, poll, and read b64_json.
The Image 2 flow is asynchronous: submit with gpt-image-2, poll the task, then read the returned data array.
curl https://api.alltoken.ai/v1/images/generations/async \ -H "Authorization: Bearer $ALLTOKEN_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $REQUEST_ID" \ -d '{ "model": "gpt-image-2", "prompt": "A premium skincare bottle on frosted glass with soft morning light, clean product photography, subtle shadows, 1:1.", "size": "1024x1024", "quality": "high", "output_format": "png"}'curl https://api.alltoken.ai/v1/images/generations/$TASK_ID \ -H "Authorization: Bearer $ALLTOKEN_API_KEY"Showcase
Six starter directions for Image 2.6 examples
Six AllToken-generated examples spanning product, editorial, landscape, portrait, abstract material, and typography.
A premium skincare bottle on frosted glass with soft morning light, clean product photography, subtle shadows, 1:1.
A friendly editorial illustration of a robotics workshop, warm color palette, clean shapes, detailed but readable, 1:1.
A wide alpine lake at sunrise with glassy reflections and low mist, photorealistic landscape, 3:2.
A cinematic studio portrait with natural skin texture, soft rim light, neutral background, professional editorial look, 2:3.
Abstract translucent ribbons folding through a dark studio space, luminous edges, high-detail material study, 1:1.
A crisp poster mockup that reads "IMAGE 2" in bold type, precise typography, paper texture, warm accent color, 1:1.
Image to image
Edit, mask, and vary - start from an image.
Image 2 also takes an image as input: restyle the whole frame, change a masked region only, or generate variations from a source.
Keep the same vase and stem; restage into a warm sunset-lit studio with a soft amber-to-ivory gradient backdrop and a subtle long shadow.
Within the masked sky only, replace with a dramatic golden-hour sunset and soft layered clouds; leave cliffs and sea unchanged.
No prompt - a variation generated from the source.
Pricing
Pricing is shown explicitly on the model detail page and in the dashboard before use.
Size options1024x1024 / 1536x1024 / 1024x1536
Quality optionsauto / low / medium / high
Price referenceLoading pricing…
See full pricing ->FAQ
Image 2 FAQ
What is Image 2?
Image 2 is AllToken's dedicated surface for OpenAI flagship image generation through the gpt-image-2 model.
How do I start?
Create an AllToken API Key, use the AllToken base URL, submit an async image task, then poll until the response returns a data array.
How is GPT Image 2 different from DALL-E 3?
gpt-image-2 is positioned as the current flagship image model surface in this workflow, while DALL-E 3 remains a separate OpenAI image model family.
How is Image 2 billed?
Image generation usage is billed according to the pricing shown on the model detail page and in the dashboard before use. Customers can review applicable usage rates before sending requests.
Can I use outputs commercially?
Commercial use depends on your AllToken Terms and upstream OpenAI policy obligations. Review both before shipping generated assets.
Studio is live