Art Prompt Generator
Pick a style, a subject, a mood and a model, or choose surprise me, then press Generate. You get a ready-to-paste art prompt in seconds. Press the button again for a new idea, or show five at once. It runs in your browser and needs no account.
Your prompt
Your prompt will appear here.
Options: style, subject, mood, model
- Style: watercolour, oil painting, digital painting, ink drawing, paper cut, pixel art, surreal or anime illustration.
- Subject: a character, landscape, animal, city scene, still life or fantasy creature. Each list holds several ideas, and the tool picks one at random.
- Mood: calm, dramatic, playful, mysterious or warm and cosy. Each mood adds light and colour words.
- Model: Midjourney gets short phrases and parameters, Stable Diffusion gets a negative prompt, and the rest get clear sentences.
Choose "Surprise me" for any option to let the tool decide.
Example outputs
These five were produced by the tool itself with the options shown.
Style: watercolor, subject: landscape, mood: calm, model: midjourney
A snowy mountain pass with a winding path, soft watercolour painting with loose washes and visible paper texture, quiet and peaceful, soft morning light, pale colours, highly detailed --ar 4:5 --v 7
Style: oil, subject: character, mood: dramatic, model: nano-banana
An old fisherman mending a net. Style: oil painting with thick brush strokes and rich colour. Mood: dramatic and intense, with strong contrast, deep shadows, stormy sky. No text and no watermark.
Style: papercut, subject: city, mood: playful, model: flux
An old harbour town at sunrise. Style: layered paper cut artwork with soft shadows between the layers. Mood: playful and bright, with saturated colours, whimsical details. No text and no watermark.
Style: ink, subject: animal, mood: mysterious, model: stable-diffusion
A heron standing in shallow water at dawn, black ink drawing with fine cross hatching on cream paper, mysterious and moody, fog, low light, hidden details, masterpiece, best quality Negative prompt: blurry, lowres, watermark, text
Style: digital, subject: fantasy, mood: warm, model: chatgpt
A floating island with a tiny village. Style: detailed digital painting with clean edges and glowing highlights. Mood: warm and cosy, with amber light, soft textures, gentle glow. No text and no watermark.
How to use the result in Gemini and ChatGPT
Gemini. Open a chat in the Gemini app, upload your photo with the plus button if the prompt mentions it, paste the result and send. Ask for one change at a time.
ChatGPT. Start a new chat, attach your photo if needed, paste the result and ask for an image. If ChatGPT replies with words only, say "create the image now".
Frequently asked questions
Are the prompts really random?
The tool picks a subject at random from a fixed list for each option you leave on surprise me, so you will see repeats over time.
Can I use the prompts commercially?
The prompts are short ideas you may use freely. Check the terms of the AI tool you paste them into for the images it makes.
Why do the results differ between models?
Each model reads words differently. Try the same prompt in two models and keep the one you prefer.
Can I change the prompt after generating it?
Yes. Copy it and edit any word. Small changes make big differences.