The Image Generation API built for scale
Nano Banana, GPT Image, Flux and more in one API. Every image is stored, transformed and delivered by Cloudinary.
Trusted by 13K brands and 4M developers
From prompt to production image in one call
An image generation API turns a text prompt, and optionally a reference image, into a new image from your own code. Cloudinary’s Image Generation API gives you 10 models, including Nano Banana, GPT Image and Flux, and stores every result in your media library, ready to transform and deliver.
- Generate with the model that fits the job, chosen with one parameter.
- Keep every image as a normal Cloudinary asset, not a temporary file.
- Deliver it resized and in the best format from a global CDN with
f_auto,q_auto.
Generation is step one. Cloudinary runs the whole pipeline.
Other tools hand you a file. Cloudinary keeps every generated image in your media library, so you can transform, tag and deliver it with the APIs you already use.
Generate
Choose from 10 models with one parameter.
Store
Images land in your library as normal assets.
Tag
Auto-tag so teams and agents find them later.
Optimize
Crop, resize and convert with f_auto,q_auto.
Deliver
Serve every device from a global CDN.
For teams generating thousands of images, not one
Product catalogs
Consistent shots and variants for every SKU, delivered to your store and marketplaces.
Ad creative variants
On-brand versions for every campaign, size and channel from one source image.
In-app generation
Let your users generate images in your product, stored and served by Cloudinary.
Free tools, demos and docs
AI Image Generator
Try prompts in your browser. No code needed.
Try it free → DocsMCP server
Let Cursor, Claude and other agents generate and manage images.
Set it up → DocsGenerative AI transformations
Background replace, generative fill and more, from the URL.
Read the docs → GuideImage Generation API guide
How generation APIs work and what to look for.
Read the guide →Image generation API questions
What is an image generation API?
An image generation API turns a text prompt, and optionally a reference image, into a new image from code. Cloudinary’s Image Generation API also stores each result in your media library, so you can transform and deliver it like any other asset.
Which models does Cloudinary’s Image Generation API support?
It gives you 10 models, including Nano Banana, GPT Image and Flux. You pick the model with one parameter, so you can switch without changing how you store or deliver images.
How is this different from calling a model provider directly?
A model provider returns a file. Cloudinary generates, stores, tags and transforms the image in one workflow, then delivers it from a global CDN in the best format for each device.
Can AI coding agents use it?
Yes. With the Cloudinary MCP server, agents in tools like Cursor and Claude can generate, upload, transform and manage images for you.
How do I generate images at scale?
Call the API from your backend or a batch job, write results to a folder per product or campaign, and serve them through URL transformations such as f_auto,q_auto, so every image is sized and compressed automatically.
Is there a free plan?
Yes. You can sign up for free with no credit card. Image generation uses your plan’s credits.
More ways to use image generation
Generate your first image in minutes
Sign up for free. No credit card required.
