AIモデル · editing

Krea 3

Fast image editing and text-to-image generation with plain-language control.

10 CR / RUN · editing · 23.7K RUNS

Krea 3

概要

Krea 3 is built for fast image editing with simple, natural-language instructions. You can guide changes directly in plain English instead of relying on complex workflows.

It also generates images from text, making it useful for both creating visuals from scratch and refining existing images. With options like aspect ratio, resolution, negative prompts, and regional edits, it gives you flexible control over the final result.

強み

  • 01Fast image editing from plain-language prompts
  • 02Supports both editing and text-to-image generation
  • 03Regional edit controls for targeted changes
  • 04Negative prompts for tighter creative direction
  • 05Adjustable aspect ratio and resolution settings

プロンプト例

Edit this product photo to add soft studio lighting, clean reflections, and a premium minimal background.

このプロンプトを試す

Generate a cinematic portrait of a woman in a red coat walking through neon-lit rain at night, highly detailed.

このプロンプトを試す

Change only the sky in this landscape to a dramatic golden-hour sunset while keeping the mountains untouched.

このプロンプトを試す

Create a cozy modern living room with warm wood tones, large windows, and Scandinavian styling, 16:9.

このプロンプトを試す

活用例

Quick image touch-ups

Use plain-language prompts to refine existing visuals without a complicated editing process. It's a practical fit for fast creative iterations.

Targeted regional edits

Make focused changes to specific parts of an image instead of reworking the whole composition. This is useful when only one area needs adjustment.

Text-to-image concept creation

Generate new images from a written prompt when you need fresh visual directions. It's helpful for exploring ideas before moving into production.

Format-ready creative outputs

Set aspect ratio and resolution to shape outputs for different layouts and channels. That makes it easier to create images that match your intended format.

よくある質問

関連モデル