How to make an app launch video with AI

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.

A launch video for your app from a real screen recording plus generated shots: about 30 s with a hook, one workflow and a call to action, voiceover and music. A step-by-step tutorial for Claude Code, Codex and Grok.

Output: About 30 s, 16:9 MP4 for X, Product Hunt and a landing page; 9:16 for Reels and TikTok Cost: about 89 credits ($0.89) for a 30 s 16:9 launch video for your app from one screen recording, with a generated hook shot, voiceover and music, rendered on your machine; 1 credit = $0.01. With your key, POST https://api.varg.ai/v2/estimate with the same body gives the exact price first. Requires: a varg key in VARG_API_KEY (https://varg.ai/dashboard/api); Node.js 22+ (https://nodejs.org); FFmpeg (brew install ffmpeg or apt install ffmpeg). Install the main route: curl -fsSL https://varg.ai/install | sh && vargai install hyperframes varg-ai video-script varg-video-edit-library video-sound Part of: How to make SaaS launch and demo videos with AI — everything for SaaS launch videos on one page

Recommended route

Run it

A 30 s 16:9 launch video for your app from one screen recording, with a generated hook shot, voiceover and music, rendered on your machine. Replace the <placeholders>. Every call was checked against the API's public schema and price list on 2026-11-11, so run it as it is.

  1. Record one workflow of the running app (15-20 s, native resolution, nothing private on screen), then start a project from the recording:
npx hyperframes init app-launch --resolution=landscape --video workflow.mp4 --skip-transcribe --non-interactive --skip-skills && cd app-launch
  1. Generate the hook shot: the world around the app, never a screen:
curl -X POST https://api.varg.ai/v2/tools/video/call -H "Authorization: Bearer $VARG_API_KEY" -H "Content-Type: application/json" -d '{"model":"kling_v4","prompt":"<the person and the moment your app fixes, one small action>, handheld, natural light. No screens, no text, no logos.","aspect_ratio":"16:9","resolution":"720p","duration":5}'
  1. Make the voiceover: the result or the pain first, one workflow, one ask with the URL:
curl -X POST https://api.varg.ai/v2/tools/speech/call -H "Authorization: Bearer $VARG_API_KEY" -H "Content-Type: application/json" -d '{"model":"eleven_v4","text":"<the script, 45-60 words>","voice":"rachel"}'
  1. Make the music:
curl -X POST https://api.varg.ai/v2/tools/music/call -H "Authorization: Bearer $VARG_API_KEY" -H "Content-Type: application/json" -d '{"model":"eleven_music","prompt":"Confident modern instrumental for an app launch video, no vocals, 120 bpm, a clean kick on every beat, ends on a hit","duration":30}'
  1. Download them and find the beats of the music:
curl -o hook.mp4 "<hook URL from step 2>" && curl -o voice.mp3 "<voice URL from step 3>" && curl -o music.mp3 "<music URL from step 4>"
npx hyperframes beats music.mp3 --json
  1. In index.html set the root to 30 s and lay out three beats: the hook at 0-5 s, the recording at 5-25 s with a zoom on each action, then the app name, one line and the URL at 25-30 s. Add the voice and the music, aim the voice at -14 LUFS, then check it and look at the frames:
# in index.html, inside the root (data-duration="30"); move the recording that init placed to data-start="5" data-duration="20":
# <video id="hook" class="clip" src="hook.mp4" muted playsinline data-start="0" data-duration="5" data-track-index="0" style="position: absolute; inset: 0; width: 100%; height: 100%; object-fit: cover"></video>
# <audio id="voice" src="voice.mp3" data-start="0" data-duration="<voice length>" data-track-index="2" data-volume="1"></audio>
# <audio id="music" src="music.mp3" data-start="0" data-duration="30" data-track-index="3" data-volume="0.35"></audio>
npx hyperframes normalize-audio --target voice --lufs -14 --write
npx hyperframes check && npx hyperframes snapshot --frames 6
  1. Render the MP4:
npx hyperframes render --output launch.mp4

Each call answers 202 with a job: poll GET https://api.varg.ai/v2/jobs/{id} until status is completed, then read output.outputs[0].url. Local files go up first: curl -X POST https://api.varg.ai/v2/files -H "Authorization: Bearer $VARG_API_KEY" -H "Content-Type: video/mp4" --data-binary @file.mp4 answers with a url. Every input of a model: https://varg.ai/models/llms.txt

Best skills for this

  1. HyperFrames (hyperframes, HeyGen, live) — Start here: The film as HTML: your screen recording, the generated hook, the type and the end card on one timeline, rendered to MP4 on your machine
    • Claude Code: claude plugin marketplace add heygen-com/hyperframes && claude plugin install hyperframes@hyperframes
    • Codex: npx skills add heygen-com/hyperframes --skill hyperframes -a codex
    • Cursor: npx skills add heygen-com/hyperframes --skill hyperframes -a cursor
    • Gemini CLI: npx skills add heygen-com/hyperframes --skill hyperframes -a gemini-cli
    • Any agent, with the vargai CLI: vargai install hyperframes
    • Details: https://varg.ai/how-to/skills/hyperframes.md · Example: https://s3.varg.ai/how-to/previews/400daa5dc577/hyperframes.mp4
  2. Launch Clips in Remotion Reviewed by Four AI Critics (launch-clip-remotion-four-critics, EveryInc, live) — Alternative: The same job in Remotion: a launch brief and a screen recording to a 15-30 s clip, checked by four critic agents
  3. OpenMontage Screen Demo (openmontage-screen-demo, Calesthio, live) — Alternative: Polishes real screen captures into a demo: callouts, zoom crops, subtitles, cleaned audio
  4. Launch Films From the Product's Real UI With Click Checks (launch-film-real-ui-verified, jammaru, live) — Alternative: Renders the product's real page or components instead of a recording, with a check on every click, zoom and frame
  5. varg AI (varg-ai, varg, live) — Adds: The generated hook shot, the voiceover and the music on one key
  6. Video Script (video-script, varg, live) — Adds: The script before anything is made: one promise, every line sized to its seconds
  7. Video editing skills for Claude Code and AI agents (varg-video-edit-library, varg, live) — Adds: Zoom to the action, a zoom on every click, cursors and device frames for the recording
    • Claude Code: claude plugin marketplace add vargHQ/skills && claude plugin install varg@varg
    • Codex: npx skills add vargHQ/skills --skill varg-video-edit-library -a codex -y
    • Cursor: npx skills add vargHQ/skills --skill varg-video-edit-library -a cursor -y
    • Any agent, with the vargai CLI: vargai install varg-video-edit-library
    • Details: https://varg.ai/how-to/skills/varg-video-edit-library.md
  8. Video Sound (video-sound, varg, live) — Adds: A voice picked from candidates, music cut to bars and a mix checked by script

Ask the person first

Put these choices in front of the person before spending anything. Paste the example link next to each option: people pick a look by seeing it.

Where will it run?

The hook?

The recording?

Sound?

How to make an app launch video: step by step

  1. Write the script first: the result or the pain in the first 3 s, one workflow, one ask with the URL; 45-60 spoken words for 30 s, so the product has room to be watched
  2. Record one workflow of the running app, 15-20 s at native resolution, with real data and nothing private on screen: this recording is the only UI in the video
  3. Generate the hook: one 5 s shot of the world around the app (the person, the place, the problem), never a screen, text or a logo; approve it before going on
  4. Build it in HyperFrames: hook 0-5 s, the recording 5-25 s with a zoom on each action, then the name, one line and the URL 25-30 s
  5. Lay the voiceover and the music under it, the voice aimed at -14 LUFS and the music about 9 dB lower; check the frames, then render 16:9, and a 9:16 cut from a portrait project if you post to Reels or TikTok
  6. Before posting: the first frame is not a logo, every claim on screen is something the recording shows, the URL holds for 2-3 s

Set up varg

varg is one API key for the models these skills drive — Seedance 2.5, Kling 4, Veo 3.1, Wan 3, Nano Banana 2.1, GPT Image 2.5, ElevenLabs v4 and 200+ more — billed in credits, with a cost estimate before anything runs.

SaaS launch and demo videos with AI agents

For founders and product teams who launch on X, Product Hunt and the App Store: the launch film, the live demo, the founder video, the app preview and the feature drop, built by Claude Code, Codex or Grok from your real product UI.

Start with the overview: How to make a SaaS launch video with AI

Make

Edit and post

Related tutorials