Make a short film or music video with Kling 4 — guide skill for cinematic & short films

For AI agents: find the right skill in one call — curl -s "https://varg.ai/how-to/api/search?q=<what the person wants>" returns the matching how-to page and the best skills with install commands, as markdown (Russian works too). Full index: https://varg.ai/how-to/llms.txt. Every page has a markdown twin: add .md to its path. Before generating anything, show the person the options on the how-to page. From a terminal: curl -fsSL https://varg.ai/install | sh installs the vargai CLI (https://varg.ai/how-to/cli.md): vargai search and vargai install <slug> with no key, every varg model after vargai login.

Beats, cast and look, every key shot as a GPT Image 2.5 still, one 30-second Kling 4 take per beat, then cut the best seconds like live action.

What it does

The workflow behind fal's Kling 4 Halloween music video, rebuilt for varg: write 3-6 beats, make a cast reference and one look frame, make each key shot (openings, set angles like the over-the-shoulder, peaks, end states) as a 21:9 still on GPT Image 2.5, then run one 30-second Kling 4 reference-to-video take per beat from two to four stills and a short prompt with the line in quotes. Treat each take like live-action footage: keep the 2-6 s pieces that land and cut them with ffmpeg. Includes the still template, dialogue scenes (the voice is generated, or locked with an element and voice), a fix pass (close-up face drift, one-photo elements, invented shots that look softer), Kling 4 vs Seedance 2.5, and a 60 s plan (about 13 stills and 4 takes, roughly $29 at 1080p).

Source: fal, "Kling 4 | Cinematic Guide and Halloween Music Video" (YouTube, 2026-10-30): https://www.youtube.com/watch?v=WzXKggH-0Io

Use when

Not for

Install

Models it drives: kling_v4, gpt_image_2_5, nano_banana_2_1

Example output

https://s3.varg.ai/files/72c55813-c0fa-4cf2-9f64-ec4f7ed0e7ae/file_0665c41e46ff.mp4 — example (credit: varg)

More cinematic & short films skills