About geminiprompts.world
AI generative media moves at blistering speed. Valuable prompting techniques launch on social networks, diffuse across reposts, and quickly disappear. We curate the prompts that genuinely work — verified by the exact clip or image they produced — and structure them so creators can filter by camera movement, lighting, mood, or model.
Watch before you copy
Every prompt entry links to the verified clip or image it created. No blind guesswork — you inspect the visual outcome before borrowing the prompt.
Searchable taxonomy
Camera moves, lighting styles, moods, duration, aspect ratios, and difficulty levels are tagged consistently and filterable instantly in the library.
Updated daily
Fresh entries are reviewed and cataloged daily as new generative checkpoints, weights, and prompting workflows emerge across the AI ecosystem.
How entries are made
Building a high-signal prompt repository requires strict editorial standards. We do not automatically scrape or dump unverified text. Every entry in the library goes through a consistent evaluation process:
- Selection Criteria: We look for prompts that showcase non-trivial cinematic technique, creative composition, or notable model capabilities. We specifically prioritize prompts that use precise cinematography terminology (e.g., tracking shots, rack focus, golden hour lighting, volumetric haze). We exclude spam, low-effort keyword stuffing, self-promotional text, and broken links.
- Verification Workflow: Every candidate prompt must be accompanied by verified visual evidence. We verify that the generated video clip or image directly corresponds to the prompt text, verifying that key parameters (camera movement, lighting conditions, subjects) are visibly represented in the resulting media. If a prompt cannot be verified with working visual output, it is rejected.
- Update Frequency: Our collection is updated daily. As frontier video-generation and image-generation models receive architectural and prompt-adherence updates, we continuously index new examples and update taxonomy metrics.
Who runs this site
geminiprompts.world is an independent benchmark and educational archive owned and maintained by [Owner Name], based in [Location].
For questions, creator attributions, collaborations, or removal requests, you can reach the operator directly via email at vibecutz123@gmail.com or visit our dedicated Contact page.
Models covered
The library dynamically indexes prompts across frontier AI video and image models in our catalog:
| Model | Maker | Verified Prompts |
|---|---|---|
| Veo | 15 | |
| Seedance | ByteDance | 1,056 |
| Grok Imagine | xAI | 1,959 |
| Gemini Image | 11 | |
| Total Catalog | All Makers | 3,041 (3,029 video / 12 image) |
Creator rights & ethics
- No Re-Hosting: Video and image media are hotlinked from original public posts on X's CDN (twimg.com). We do not download, store, or re-host original video files on our servers.
- Full Attribution: Every prompt explicitly links to the original creator's public profile and post.
- Fast Removal Policy: Any author can request removal of their prompt or media link at any time with no questions asked via our Removal Requests form.