Trending
Prompts
How the AI prompt catalog works
The mozgi.io catalog has more than 33,000 ready-made prompts for generating images and video. Each card includes a sample result, the full prompt text, the model, the aspect ratio, the author and the source. The prompts are collected from open sources: X (Twitter), GitHub, and authors' blogs and articles. Some of the prompts are written by us, and their author is @mozgi.io. New prompts are added every day.
Which AI models are in the catalog
For images: GPT Image 2.0 and 2.5, Nano Banana Pro, Nano Banana 2 and 2.1, Seedream 4.5 / 5.0. For video: Seedance 2.0 and 2.5, Grok Imagine Video, Gemini Video, plus the first prompts for Kling, Wan and Minimax. GPT Image 2.0, Nano Banana Pro and Seedance 2.0 have the most examples. The "Code" section collects prompts for Claude Opus 5.5: with them, the model writes code for interfaces, mini-games and animations.
Who it's for
- Designers: posters, mockups, infographics, illustrations.
- Social media managers: images and short videos for social networks.
- Online sellers and marketers: product photography, ad scenes with products, infographics for product listings.
- Creators: anime, fantasy, cinematic scenes, storyboards for video.
How to use the catalog
- Find a prompt using search, the model filter (images, video, code), a genre, or sorting by likes and date.
- Open the card: it shows a sample result and the full text.
- Click "Copy".
- Add your own details. Words in [square brackets] mark the places for your data: the subject, a color, a caption, a description of your photo. If there are no brackets, replace the subject, location and details right in the text.
- Paste the prompt into a service with the right model and start the generation.
How to choose a model
Start with the model listed on the card. The author wrote the prompt for it, so that model will give you the result closest to the example. A prompt usually works in another model of the same type too, but the lighting, composition and text in the image may come out differently. Video prompts describe camera movement and action over time, so they are meant for video models: Seedance, Grok Imagine Video, Gemini Video. If you don't know where to start, pick a model with many examples.
What the genres mean
A genre shows the task a prompt was written for. "Avatar / Portrait" and "Photo studio" cover portraits and studio shoots. "Hero shot" is a frame built around one person. "Product marketing" and "Infographics" cover ad scenes with products, diagrams and cards with text. "Social media" and "Poster" cover visuals for posts and event posters. "Cinematic", "Sports" and "Spaces" cover scenes in motion, indoors, on the street and in nature. "Anime", "Fantasy" and "Drawing" are about stylization. "Storyboard" covers comics and frame sequences. "Design" covers interfaces, mockups and logos.
Frequently asked questions
Are the prompts free?
Yes. You can view and copy prompts for free and without signing up. Generation is paid for in the AI service at its own rates: mozgi.io does not generate images or video and does not sell generations.
Can I use the prompts and results commercially?
Prompts belong to their authors: each card lists the author and the source. Prompts by other authors are published for reference and personal use. If you publish a prompt or a result, credit the author and link to the original. You are responsible for any commercial use. Original @mozgi.io prompts can be used freely for yourself and your projects; commercial use is allowed only with written permission, and republishing requires a link to mozgi.io. Rights to the image or video itself are defined by the terms of the service where you generated it. Learn more in the terms of use.
How do I use my own photo?
If a prompt mentions a reference (reference image, "from the photo", [reference photo]), upload your image to the generation service together with the prompt. This is how models that accept images as input work: Nano Banana and GPT Image edit the uploaded photo, and Seedance makes a video from an image. Check your service's settings to see which files it accepts.
Why does my result differ from the example?
The AI generates the image from scratch every time, so even with the same prompt the result will be different. It depends on the model version, the settings (aspect ratio, quality), the service itself and the source images. Set the aspect ratio as shown on the card, run several generations and spell out the details that matter to you in the prompt.
What language should I write prompts in?
Most prompts in the catalog are in English: that's the language their authors published them in. The example on the card was made with exactly this text, so without translation your result will be closer to the example. Many models understand other languages as well. If you need text in the image, write it in the language you want, in quotes.
Are prompts available in other languages?
Most prompts are in the author's language, usually English. Some original @mozgi.io prompts (author @mozgi.io) are written in Russian. Card titles and descriptions in the catalog are available in English.
How often is the catalog updated?
Every day. New prompts are shown first by default. You can get a selection of new prompts in the Telegram channel t.me/mozgi_io.