Image generation

**Krea Image **is the platform’s core creation tool. From a single text prompt, you can generate everything from photorealistic photographs to stylized illustrations, concept art, and graphic design assets. The tool gives you access to multiple AI models, each with different strengths, all from one interface.

​

How it works

Open image mode using the image icon in the left navigation bar. You will see a prompt box at the bottom of the screen and a model selector in the bottom-left corner.

​

1. Choose a model

Click the model name in the bottom-left to open the model picker. The picker lists all available image models along with a short description of each one, its speed, quality tier, and compute cost. The default selection is Auto, which chooses the most suitable model for your prompt automatically. If you prefer to choose yourself, a good place to start is Krea 1, Krea’s own frontier model, which is optimized for stylish, photographic-style images that avoid a generic AI look. The full list of available models is covered in the Model Overview. Images per generation: Krea 1, Flux, and Flux.1 Krea generate four images per prompt. All other models generate two. Compute units per generation: Each paid plan includes a default compute unit budget. Models vary significantly in cost — a Krea 1 generation uses 8 units, while ChatGPT Image uses 183. Iterate with lower-cost models where possible, and purchase add-on units if needed. Speed: Generation times range from around 6 seconds (Krea 1) to 45 seconds or more (Nano Banana). Resolution: Most models cap at 1024px square. Krea 1, Seedream 4, Wan 2.5, Flux.1 Krea, and Flux 1.1 Pro Ultra support higher-resolution outputs, up to 4K in some cases. Here is a full breakdown: Models also vary by what other settings they offer. For example, only some models accept Styles. Here’s a breakdown: *Note: Seedream 4, Nano Banana and ChatGPT Image are reasoning models with intelligent understanding of prompts. Although they do not offer ‘character reference’ or ‘style reference’ as explicit settings, you can upload images of characters or artwork styles, and simply write a prompt that says "Create an illustration of a skyscraper in the style of the attached image."

​

2. Write a prompt

Describe the image you want in the text box. You can be as brief or as detailed as you like. For most models, including subject, setting, lighting, and mood in your description will produce more precise results.

​

3. Adjust settings

Most models offer additional settings below the prompt box. These let you control aspect ratio, apply a style, upload a reference image, and more. Not every setting is available for every model. The full list of settings is explained below.

​

4. Generate

Click Generate. Depending on the model, generation takes between 4 and 180 seconds. Most generations return either two or four images per prompt.

---

​

The model picker

The list is ordered with the most recent and popular models at the top. To see the full range of available models, see our Model Overview.

​

Auto mode

The Auto option at the top of the list selects the best model for your prompt and plan automatically. It is a good default if you are not sure which model to use.

​

Available models

The models shown in the picker fall into two groups. Krea models are developed and maintained by Krea: Third-party models are provided by external AI labs and are accessible directly from the same interface: For a full breakdown of every model including compute costs, generation speed, resolution limits, and supported settings, see the Model Overview.

---

​

Settings

When a model is selected, the available settings appear beneath the prompt box. Settings vary by model.

​

Aspect ratio

Controls the dimensions of the output image. Most models offer preset ratios (square, portrait, landscape) and some support custom ratios. Use 9:16 for vertical content such as social media stories or TikTok.

​

Style transfer

Available for: Krea 1 only. For other models, use the Image prompt setting with a similar instruction in your text prompt. Transfers the visual look of an uploaded image onto your generation. Upload your own image, choose from Krea’s provided starters, or reference a previously generated image. Useful for matching a consistent aesthetic across multiple outputs.

​

Image prompt

Upload one or more images to influence your generation. In most models, elements of the uploaded image, such as composition, color, and style, are blended with your text prompt. In reasoning models such as Nano Banana, you can reference the uploaded image directly in your prompt (for example, ‘Create an editorial photo of the woman in the attached image, wearing the red dress in the other image’). The number of images you can upload varies by model, from 2 in Krea 1 to up to 15 in Nano Banana.

​

After generating

Each prompt produces a row of results. Hover over any row to reveal additional controls. Below the images:

• Retry re-runs the same prompt and model immediately.
• Reuse parameters loads the prompt, model, and settings back into the prompt box for editing.
• Share parameters generates a link that others can use to reproduce your generation exactly.
• Download all downloads the full set of images as a .zip file.
• Publish creates a shareable public page for your images.
• Delete removes the images from your session view. Your assets are not deleted.
  

Floating above each image on hover:

• Fullscreen shows your generated image at fullscreen
• Edit send the image to the Edit tool
• Upscale & enhance sends the image to the Enhance tool
• Video sends the image to the Video tool as a starting frame.
• Download icon saves the image to your device.
• Heart icon marks the image as a favorite for easy retrieval in Assets.

​

Sessions

Each generation is added to the current session. You can continue adjusting your prompt and generating new rows within the same session to iterate on a direction. Start a new session by clicking the + icon in the left-hand panel to keep different projects organized. To find images from past sessions, click the Assets icon in the top navigation bar to browse all your generated images in a single grid.

---

​

Choosing a model

A quick reference for common use cases: For a full model comparison including compute costs, generation speed, and supported settings, see the Model Overview.