Viral Video Generator

Viral Video Generator is an end-to-end AI video production pipeline that turns a single text prompt into a complete, ready-to-post vertical video. It generates video clips with Google Veo 3.1, synthesizes a voiceover with Piper TTS, adds background music and sound effects, and assembles everything with ffmpeg into a finished MP4 for TikTok, Instagram Reels, and YouTube Shorts.

What Is This?

• A full production pipeline, not just an idea generator: prompt in, finished vertical video out
• Combines four stages — script writing, parallel asset generation (video clips, voiceover, music/SFX), and ffmpeg assembly
• Video clips are generated with Google Veo 3.1; narration is synthesized locally with the Piper neural TTS engine
• Output is a standard MP4 formatted for short-form platforms like TikTok, Instagram Reels, and YouTube Shorts

Why Use It?

• Skip the whole editing suite — no timeline, no cutting, no manual voiceover recording
• One prompt, one video — describe the video you want and receive an assembled file, not a to-do list
• Structured for retention — scripts follow a proven 6-8 scene arc (hook, discovery, demo, features, payoff, call to action)
• Runs headless — Piper TTS and a static ffmpeg build install automatically on first run, so there is nothing to configure

How to Use It?

Describe the video you want, including topic and style, and the pipeline handles scripting, generation, and assembly:

Generate a 30-second vertical video: a capybara barista pours latte art in a cozy cafe, cinematic slow motion, warm light, meme style

You can steer tone and structure in the prompt:

Create a product promo video for an eco-friendly water bottle, upbeat narrator, 6 scenes, ending with a call to action

Behind the scenes the skill writes a scene-by-scene script (visual prompt, voiceover line, and SFX cue per scene), generates every asset in parallel, then stitches the final cut with ffmpeg.

When to Use It?

• You need short-form social content fast: product promos, meme videos, tutorials, announcements
• You want a narrated video without recording your own voice or sourcing music
• You are prototyping video concepts and want a watchable draft in minutes, not hours
• You do not need it for long-form or highly art-directed work — it is built for short, punchy vertical clips

Important Notes

• On first run the skill auto-installs Piper TTS and a static ffmpeg build; video generation runs through Happycapy's hosted models, so no local GPU is needed
• Generation is asynchronous: each scene's clip takes about 1-2 minutes, so a full 6-8 scene video takes several minutes end to end
• Output is short-form vertical video by design; for long-form or frame-precise edits, export the assets and finish in a video editor
• In our verification the core generation step (Veo-class clip via the platform gateway) was executed for real; the TTS/music/assembly stages were reviewed in the skill source

Try It in Happycapy

1. Open Happycapy in your browser (no installation or signup needed to try)
2. Ask for the video directly: "Make me a 30-second vertical video about a robot learning to cook, funny style"
3. Wait for the pipeline to generate clips, voiceover, and music, then download the finished MP4