Craiyon AI Image Creator

🎨 Craiyon AI Image Creator for text prompts

Craiyon AI Image Creator generates images from text prompts on Craiyon. Add one or more prompts, choose a Craiyon style and aspect ratio, then export Apify download links, Craiyon source URLs, dimensions, categories, and generated timestamps from a clean Apify dataset.

Use this Actor when you want repeatable Craiyon text-to-image output for mood boards, concept art, icon ideas, product mockups, creative briefs, content drafts, or prompt experiments. Instead of opening Craiyon by hand for each idea, you can run a small batch in Apify Console, call the Actor through the Apify API, schedule runs, and export the results as JSON, CSV, Excel, XML, RSS, or HTML.

For a quick first run, keep the sample prompt, leave Images per prompt set to 1, and keep the default automatic style and aspect ratio. Once the output shape looks right, add more prompts or adjust the style, aspect ratio, and optional negative prompt.

✅ What this Actor does

• Generates Craiyon images from one or more text prompts.
• Saves one dataset row per successfully generated image.
• Saves every accepted image and returns an Apify download URL.
• Returns the direct source image URL from Craiyon when available.
• Supports Craiyon style choices: automatic, artistic, photo, illustration, anime, vector, and raw.
• Supports automatic sizing, square, landscape, and portrait requests, and reports the actual generated shape in each output row.
• Lets you add a negative prompt when you want Craiyon to avoid specific visual elements.
• Includes dimensions, categories, resolved style, actual generated shape, and generated time.
• Skips failure and status placeholder rows so the dataset contains generated images only.

The Actor is built for Craiyon's public prompt generation flow. You do not need to provide a Craiyon account, cookies, or an API key.

📦 Data you can export

Each output item represents one successfully generated Craiyon image. Fields include:

• prompt and promptIndex
• imageIndex
• sourceImageUrl
• width and height
• category
• style
• aspectRatio
• generatedAt
• downloadUrl

The sourceImageUrl points to the generated image returned by Craiyon. The downloadUrl points to the copy saved by this Actor, which is usually the safer URL to use in downstream workflows.

Rows are saved as images finish, so a multi-prompt run can appear out of input order. Use promptIndex and imageIndex to match each image back to the requested prompt and image number.

🚀 How to run

1. Add one or more image prompts in Prompts.
2. Keep Images per prompt at 1 for the first run.
3. Leave Style and Aspect ratio on automatic unless you need a specific shape.
4. Optionally add a Negative prompt such as text, watermark, blur.
5. Start the Actor.
6. Open the dataset and use downloadUrl or sourceImageUrl for each generated image.

Clear visual prompts work best. Include the subject, medium, scene, lighting, color, composition, or style you care about.

a simple teal glass button icon with soft studio lighting on a white background
a cozy reading nook with warm morning light, interior design render
a futuristic sneaker product photo on a reflective floor

⚙️ Input

{
	"prompts": [
		"a simple teal glass button icon with soft studio lighting on a white background"
	],
	"imagesPerPrompt": 1,
	"negativePrompt": "",
	"style": "auto",
	"aspectRatio": "auto"
}

✍️ Prompts

Enter text-to-image prompts. Empty and duplicate prompts are ignored. The Actor accepts up to 100 unique prompts in one run.

🖼️ Generation settings

Use Images per prompt to request 1 to 4 saved images for each prompt. Use Style to choose Craiyon's generation style. Use Aspect ratio to keep Craiyon's automatic sizing or request square, landscape, or portrait output. Craiyon may still return a different final shape, which the Actor reports in aspectRatio.

Use Negative prompt when you want to avoid specific elements such as text, watermark, blur, extra limbs, or a color you do not want in the scene.

📄 Output example

{
	"prompt": "a simple teal glass button icon with soft studio lighting on a white background",
	"promptIndex": 0,
	"imageIndex": 0,
	"downloadUrl": "https://api.apify.com/v2/key-value-stores/example-store/records/craiyon-001-001-exampleImageId.webp",
	"sourceImageUrl": "https://img.craiyon.com/2026-06-01/example.webp",
	"width": 1024,
	"height": 1024,
	"category": ["icon", "graphic design", "minimalist"],
	"style": "vector",
	"aspectRatio": "square",
	"generatedAt": "2026-06-01T11:52:54.185Z"
}

💸 Pricing

This Actor uses pay-per-event pricing. You are charged for each generated image that is successfully saved for download and export. Invalid input, skipped generations, failed downloads, and runs that do not save generated images do not create generated-image events.

There is no separate Actor-start charge in this pricing setup.

⚠️ Limits and caveats

Craiyon can sometimes reject a prompt, return no image, temporarily block anonymous generation, or be unavailable. When that happens, the Actor logs a warning, skips the failed image, and keeps any generated images already saved in the dataset.

The dataset contains only generated image rows. It does not include failed prompt rows, screenshots, source page HTML, prompt quality scores, or image editing results.

Generated media availability can depend on both Craiyon and Apify storage. Use downloadUrl when you want the copy saved by the Actor.

❓ FAQ

🔐 Do I need a Craiyon account?

No. The Actor runs against Craiyon's public prompt generation flow and does not ask you for Craiyon cookies, login details, or an API key.

⚠️ Why did a prompt return no image?

Craiyon may reject some prompts, temporarily block anonymous generation, run out of capacity, or fail to return a downloadable image. The Actor handles that as a skipped generation and continues with the rest of the run when possible.

🔗 Which URL should I use in my workflow?

Use downloadUrl when you want the image copy saved by the Actor. Use sourceImageUrl when you specifically need the original Craiyon-hosted image URL returned by the source.

🖼️ Can I generate several images for the same prompt?

Yes. Set Images per prompt from 1 to 4. Each successfully saved image creates one dataset row and one generated-image pricing event.

📝 Changelog

• 1.0: Simplified output rows to focus on image download, source URL, prompt mapping, resolved style and shape, dimensions, categories, and generated time.
• 0.1: Initial release.

🆘 Support

For issues, questions, or feature requests, file a ticket and I'll fix or implement it in less than 24h 🫡

🔗 Other actors

• Artbreeder AI Image Creator ↗ - Generate Artbreeder images from prompts with model and shape controls.
• Unsplash Image Scraper ↗ - Collect Unsplash image search results for design and content research.
• Font Detector ↗ - Detect fonts, font files, and CSS font families from public web pages.
• Pinterest Keyword Scraper ↗ - Export Pinterest autocomplete keyword ideas for creative and SEO planning.
• Website URL Crawler ↗ - Build rendered website link maps for audits and content workflows.

Made with ❤️ by Maxime Dupré