BestAIImageGen
Video benchmark prompts
Four fixed jobs for text-to-video and editor tools. Same brief every time-so Runway vs CapCut comparisons stay honest.
Protocol
- Use the brief as written (adapt only for tool UI limits).
- Cap length at ~4–8 seconds for generative tools.
- Note credits burned, watermark, and whether you can edit the export.
- Compare with the best AI video generator table.
The four briefs
1. Product hero (text-to-video)
A matte black wireless earbud case rotates slowly on a seamless white surface, soft studio softbox light, clean commercial product video, subtle reflection, no people, no text overlays
2. Cinematic establishing shot
Drone-like push-in over a misty pine forest at sunrise, volumetric light rays, gentle camera forward move, photoreal, calm mood, no text, no logo
3. Social edit brief (CapCut / Canva)
15-second vertical Recap: 3 clips of a coffee shop morning - pour, latte art, window seat. Auto captions: "Slow mornings hit different." Upbeat lo-fi. Hard cut every 2 seconds. Export 1080x1920.
4. Avatar explainer (Synthesia-class)
Script (30s): "Here are three checks before you pick an AI video generator: motion quality, free limits, and whether the export is editable. Start with the job-not the brand." Neutral presenter, office background, no background music.
Image stills pack: benchmark-prompts. Offline copy: video-benchmark-prompts-v1.txt. Cite this page if you publish a video roundup using these briefs.