Veo 3 Prompt Generator

Free prompt generator for Veo 3, Midjourney, Imagen, ChatGPT

This helps you generate those bigfoot vlog or glass-cutting ASMR videos. Small enhancements make a big difference. Ultimately, spend time learning prompt engineering (click the Guide below), but this should help newbies get started. Here's my recommended flow:

  • Frame to Video.
    • Use Midjourney or gpt-image-1 (ChatGPT) for a starting frame in Veo 3's Frame to Video
    • Don't use Ingredients to Video, Extend, or an End Frame - these downgrade the generation to Veo 2. They'll fix this eventually, but currently Frame to Video is the only mode that uses Veo 3 Quality.
    • Frame to Video will give you better results faster, than Text to Video. You can iterate on images cheaper and faster than videos; maintain character consistency via image edits; and get consistent results by giving Veo a head start.
  • Midjourney vs ChatGPT
    • Midjourney has the highest quality results; gpt-image-1 has the strongest prompt-adherence (ie, more likely to give you what you want). For advanced users I recommend Midjourney, as it takes a PhD in prompt-engineering to get what you want. For new / intermediate users, use gpt-image-1.
    • gpt-image-1 is what ChatGPT uses; but it's better to use OpenAI's Playground because there's more control. gpt-image-1 can't output 16:9, so you'll need to crop or outpaint its 4:3 output.
    • Imagen 4, when improved, will be the best due to its native Veo 3 integration and 16:9 output. Historically it's been really crap; but they have imagen-4.0-ultra-preview which should hit Veo 3 soon. Note: Imagen can't edit existing images, only create new ones.
    • All this is why I built Generate with a frame in this tool. It handles prompt-engineering, initial-frame & follow-up edits (using gpt-image-1, since Imagen can't), and outputs 16:9.
    • See Image model differences, and keep an eye on Images Leaderboard
  • Other Tools
Comments & feedback
Enhance prompt for
Image ModelsVideo ModelsFull Workflow
Generate with a frame

Create a starting frame for your clip: either a prompt for your image tool, or generate the frame directly.

Multi-shot

Generate a sequence of connected video clips from a single prompt.

Original Prompt:

Bigfoot recording himself with a selfie stick, saying a bunch of Gen Z stuff.

Veo 3 Prompt:
Medium shot of a colossal, shaggy Bigfoot with expressive, slightly bewildered eyes, holding a modern smartphone on a selfie stick in a deep, mist-shrouded ancient redwood forest at dawn. The camera is slightly shaky, mimicking a handheld vlog. The Bigfoot turns the selfie stick towards himself, adjusts the angle clumsily, and begins to speak directly into the phone, gesturing with his free hand. He briefly pans the phone to show the majestic, moss-covered trees around him, then turns it back to himself.

Bigfoot says (with a deep, rumbling, yet earnest and slightly confused tone): "What up, fam? It's your boy, Squatch, coming at you live from the vibes! No cap, this forest slaps. Seriously vibing right now. Just dropped my new crypto, hit that subscribe, smash that like, and don't forget to ring that notification bell! We're about to get wild, for real, for real!"

Audio: The subtle rustling of leaves, distant birdsong, the muffled sound of Bigfoot's heavy breathing, and a faint, whimsical, upbeat electronic music track starting subtly as he begins to speak.
Style: Found footage, vertical video format, humorous, quirky, naturalistic lighting, cinematic.
(no subtitles, no on-screen text)
to unlock:
  • ✅Better enhancer: Stronger:🪙1 | Weaker:🪙0
  • ✅Generate starting frames for character consistency
  • ✅Save your prompt history
  • ✅API access to plug into your workflow