AI Prompt Studio. Pro prompts, first try.
Prompt recipes for business owners
The Business copy tab turns a few details into copy-ready briefs for ads, cold emails, blogs, SEO, product pages, social posts, taglines and more, so you get great output from any AI on the first try. Everything is generated locally on your device; your inputs never leave this page.
How to use the AI Prompt Studio
AI Prompt Studio builds prompts for image generators, video generators and chat assistants from a short form. You describe the subject, pick options such as style, lighting and camera, and the studio joins them into one prompt in the format your chosen tool expects. It is for marketers, founders and creators who want a solid first prompt without learning each tool's syntax.
Step by step
-
Pick the kind of prompt
Click Image, Video or Business copy above the form. Each tab has its own fields. Switching tabs starts that form fresh, so copy anything you want to keep first.
-
Describe the subject
On Image, type into Subject / scene. On Video, type into Scene & action and, if you want sound, Audio / ambience (optional). Be concrete about who, what and where; if you leave the subject blank the prompt uses a generic brand photo or brand moment.
-
Choose the look
Click one or more Style chips, then set the menus: Image type, Lighting, Camera / lens, Composition, Color & mood and Aspect ratio for images, or Shot type, Camera movement, Lighting, Mood / grade and Pacing for video. A dash means leave that detail out.
-
Set the Target tool
For images, Midjourney adds --ar with your ratio plus --style raw --v 6; DALL·E 3 rewrites the prompt as plain sentences; Stable Diffusion / SDXL adds quality tags and a separate negative prompt; Flux and Generic keep a plain list. For video, Runway Gen-3 adds a camera note and the other tools get the same text.
-
Fill in a business brief
On Business copy, choose What do you need? from 12 jobs such as Facebook / Instagram ad, Cold outreach email or SEO meta description. Fill in Business / product, Target audience, Key benefit / angle, Tone, Call to action and Anything else? (optional). The result is an instruction for ChatGPT, Claude or another chat assistant.
-
Copy and paste
The prompt updates as you type. Press Copy prompt, and on SDXL also Copy negative, then use the Paste into links under the prompt to open a matching tool in a new tab. Surprise me fills the menus and chips at random; Reset clears the current tab.
Reading the result
- Image and video prompts list your choices in a fixed order, separated by commas, and image prompts add highly detailed, sharp focus, high resolution after them. DALL·E 3 is the exception: its prompt is written as short sentences instead.
- With nothing filled in, the Image tab gives: a clean, professional brand photo, photograph, highly detailed, sharp focus, high resolution --ar 1:1 --style raw --v 6.
- The negative prompt box only appears for Stable Diffusion / SDXL. Paste it into the separate negative prompt field of your tool, not into the main prompt.
- Most business briefs give the assistant a role, then the deliverable and its limits, such as 15 headlines of up to 30 characters for a Google Search ad. An empty Business / product, Target audience or Key benefit / angle shows as a placeholder in square brackets, an empty Call to action falls back to Learn more, and an empty Anything else? (optional) is left out.
Tips and common mistakes
- Aspect ratio only reaches the prompt for Midjourney and DALL·E 3. For Stable Diffusion, Flux and Generic, set the size in the tool itself.
- For Customer support reply, use the Target audience box to describe the customer's problem. For Job description, put the role title there.
- Replace every square bracket placeholder before you send a business brief, or the assistant will write generic copy around it.
- Choose two or three styles at most. Many styles at once pull the image in different directions.
Privacy and limits
The prompt is assembled in your browser from fixed text and nothing you type is sent anywhere or saved, so a reload clears the form. The Paste into links only open the other site in a new tab; your prompt goes there only when you paste it yourself.
Everything a prompt should carry.
- Build image prompts for Midjourney, DALL·E 3, Stable Diffusion / SDXL and Flux, with image-type, style, lighting, lens, composition, colour, mood and aspect-ratio controls.
- Generate video prompts for Sora, Runway Gen-3, Google Veo, Pika and Kling, shot type, camera movement, style, mood, pacing and optional audio.
- Write 12 kinds of business-copy brief, from Meta ads and Google RSAs to cold emails, LinkedIn posts, blog outlines, SEO meta, product descriptions, taglines and video scripts.
- Hit “Surprise me” to randomise style and settings instantly when you need a creative jumpstart.
- Copy a Stable Diffusion negative prompt alongside your main image prompt with one click.
- Jump straight to the tools, one-click paste links to Midjourney, Sora, ChatGPT, Claude, Runway and more.
Frequently asked questions.
Is AI Prompt Studio free?
Is it private? Are my inputs uploaded anywhere?
Does it work offline?
Which AI tools do the prompts work with?
What is a negative prompt, and when do I use it?
Do I need an account or to install anything?
Published · Last updated .
