🎬 AI Video

AI Video Prompt Builder

Craft cinematic prompts for Seedance (Higgsfield), Veo 3, Grok Aurora, Kling AI, Hailuo, Runway, Pika, Luma, Sora & more — using the exact formula each model reads best. Generates 3 ready-to-paste variants.

🎯 Select your AI Video Tool

⚡ Quick Templates

⚠️ Seedance / Higgsfield: put your most important visual here — the model weights the first 20–30 words most heavily.
Multi-shot mode — write individual action for each numbered shot
These will be embedded inline as [VFX: your effect] in the Seedance-formatted output.
Seedance / Veo 3 support native audio. Include exact wording and voice tone for best results.
⚠️ Seedance rule: explicit negative constraints are what keep the model from drifting. Be specific.

📋 Generated Prompts

3 variants ready after generation
Fill in the fields on the left, then click
⚡ Generate 3 Variants to get copy-paste-ready prompts for your chosen tool.
Generate prompts first.
Generate prompts first.
Prompt Quality Score0%

💡 Seedance / Higgsfield Prompt Formula

1️⃣Shot structure line FIRST: Always open with shot count, duration, aspect ratio — e.g. "One continuous shot, 15 seconds, 9:16." Seedance reads this before anything else.
🎯Subject + action in the first 30 words: Anchor who/what and what they're doing immediately. The model weights the opening 20–30 words most heavily.
🔢Multi-shot = number each shot: Write "Shot 1:", "Shot 2:" etc. with explicit action per shot. Build a clear escalation arc — calm → rising tension → climax → aftermath.
[VFX: …] bracket notation: Embed effects inline inside the action — e.g. [VFX: branching electric current, blue-white arcs]. Don't describe them in a separate block.
🚫Negative constraints are mandatory: "no cuts, no zoom, natural head movement" for POV shots. "no 3D, no cartoon" to force photorealism. Without these, Seedance drifts.
🎙Veo 3 / Grok Aurora native audio: Write the exact spoken line + voice tone. These models generate in-sync dialogue — don't leave audio to chance.