Features
Each work comes with the exact prompt its author used, so you can see how lighting, camera moves, wardrobe and style were described in words instead of guessing from the result.
Cards list the model behind the work, such as GPT Image, Nano Banana, Seedance, Kling, Hailuo, Veo or Midjourney, plus the style category and the aspect ratio, from 9:16 vertical clips to 16:9 frames.
Views and likes from the source post travel with every card, so you can tell which ideas actually caught attention and not just which ones look good in a thumbnail.
Sort by hot, latest, quality, views or likes, and combine the sort with type, model, style category and ratio filters to narrow 4,400+ works down to a short list.
The "Use this prompt with" button, for example "Use this prompt with Seedance 2.0", opens the studio with the prompt and the matching model already loaded.
Every work names its creator and links to the source post on X. Authors can request removal or a change of attribution through the /copyright page.
Overview
An AI prompt gallery is a searchable collection of finished AI images and videos where each result is paired with the text that produced it. Instead of starting from a blank box, you start from something that already worked for someone else, read how it was phrased and adapt it to your own subject, product or story.
NeoMundo's AI prompt gallery at /gallery is a filterable masonry grid of more than 4,400 public works collected from creators on X. It mixes AI video prompts for models like Seedance, Kling, Hailuo and Veo with AI image prompts for GPT Image, Nano Banana and Midjourney. Each card keeps the prompt, the model, the style category, the aspect ratio and the engagement numbers from the original post.
Attribution is part of the design, not an afterthought. Every work is credited to its author with a link back to the source post, and creators who want a work removed or its credit corrected can ask through /copyright. The gallery is a place to learn from and build on published prompts, while the works themselves stay with the people who made them.
The gallery also connects directly to generation. When a work was made with a model that NeoMundo offers, one button sends the prompt and that model to the studio. From there you can change the subject, the ratio or the duration, check the credit cost shown before you start, and generate your own version with any of the 13 models on the site.
How it works
Open /gallery and pick image or video first. Then narrow by model, style category and ratio, for example video, Seedance, 9:16, to see only vertical clips from that model family.
Switch between hot, latest, quality, views and likes. Hot surfaces what is getting attention now; likes and views show what has held up over time.
Open a card to read the full prompt, check the model and ratio, and follow the link to the creator's original post if you want more context.
Click "Use this prompt with" and the studio opens with the text and matching model preloaded. Edit the subject or style, review the credit cost and generate. Images usually take 10–60 seconds and videos usually 1–5 minutes.
Use cases
Browse AI video prompts sorted by likes to find camera moves, transitions and scene descriptions that already worked on X, then rebuild them for your own 9:16 clips.
Study AI image prompts by style category to learn how others describe lighting, texture and composition, and use them as a starting brief for posters, covers or character sheets.
Filter by ratio to find product and lifestyle looks in 1:1, 4:5 or 9:16, then swap in your own product and generate ad or listing visuals.
Compare the same model across dozens of works to see which phrases change the result. Reading real prompts next to real outputs is a fast way to improve your own.
FAQ
They are public AI images and videos collected from creators on X. Each one is credited to its original author and links to the source post.
Yes. Authors can request removal or a change of attribution through the /copyright page.
The gallery covers works made with GPT Image, Nano Banana, Seedance, Kling, Hailuo, Veo, Midjourney and other models. The model is shown on every card and can be used as a filter.
The studio opens with the prompt and the matching model preloaded, so you only adjust the details and generate. For works made with a model the studio does not offer, you can still read the prompt and write your own version with one of the 13 models.
The cost depends on the model, resolution and duration, and the exact number is shown before you start. For example, Nano Banana 2 images start at 14 credits and Seedance 2.0 videos start at 80. Failed generations are refunded automatically. New accounts get 50 free credits on sign-up.
Content you generate can be used in ads, stores and social media, subject to the terms of service and the model providers' policies. The gallery works themselves belong to their creators, so do not repost them as your own.
Video prompts usually describe motion, camera movement, duration and sound, while image prompts focus on composition, light and style. Use the type filter to switch between them.