Claude Opus 5.5
Anthropic’s newest model. Faster and cheaper than the last one, and the best yet at writing code you can see.
Enrollment for this workshop is closed.
Checkout didn’t start. Please try again or email info@digitalnisuperheroj.com.
LIVE WORKSHOP · FRIDAY, OCTOBER 2 · 11AM NEW YORK / 5PM BERLIN
Make video in any style. Just describe it.
In 3 live hours you learn how to turn one sentence into a finished video: animation, 3D, AI footage, voice and music. You work in Claude Code, Codex or Cursor, whichever you already have. On Friday we build one together.
Lifetime replay · 4 ready-made templates · 14-day money-back guarantee
You don't shoot video anymore. You write it.
Until recently, good video meant a camera, an editor, an animator and weeks of waiting. Or a template that looks like everyone else’s.
Now you describe what you want. Claude Opus 5.5 writes the scenes as HTML, HyperFrames renders them into a real MP4, and models on fal and ElevenLabs add images, video shots, voice and music. Every frame is code, so you change it with one sentence.
No dragging clips around a timeline. You talk to your agent, and it does the work that used to take a whole studio.
- 01A one-sentence brief
- 02Opus 5.5 → HTML
- 03fal · ElevenLabs
- 04HyperFrames → MP4
Four videos. Zero cameras.
Everything below was made with this exact workflow: an agent plus HyperFrames. In the workshop you get the source code of all four as templates.
The promo exists in Croatian, English and Japanese: new voice, new on-screen text, same code. One video becomes three markets.
Everything in this stack is days old.
Eleven v4
ElevenLabs’ most expressive voice. It laughs, whispers and shifts tone. Released the day before we put it in our promo.
HyperFrames
HeyGen’s open-source tool that turns HTML into video. Open-sourced this spring, shipping new releases almost every day.
Most people haven’t heard of this combination yet. On Friday you’ll be using it.
THE 3 HOURS
From one sentence to a finished MP4.
We walk the whole process on one real video, step by step.
A one-sentence brief
How to tell your agent what the video sells, to whom and in what tone, so you don’t get a generic result.
Scenes as code
Opus 5.5 writes the HyperFrames composition: typography, animation, 3D particles, transitions. You learn how to ask for changes.
AI shots
Photoreal images with Nano Banana Pro and video shots with Veo 3.1 on fal. The same character across scenes.
Voice and music
ElevenLabs v4 voices in English, Japanese or dozens of other languages, with emotion and laughter, plus music cut to length.
Render and formats
MP4 for the web, 9:16 for Reels and TikTok, 4:5 for the feed. One project, three outputs.
Use what you already have.
Your agent
- Claude Code
- Codex
- Cursor
Video and animation
- HyperFrames
- GSAP
- Three.js
Images and video
- Nano Banana Pro
- Veo 3.1
- GPT Image 2.5
- fal
Voice and music
- ElevenLabs v4
- ElevenLabs Music
One agent is enough. I demo in Claude Code; you get the same instructions for Codex and Cursor.
WHAT YOU CAN MAKE
Videos you’ll know how to make
- Promos and product launches
- Meta and TikTok ads (9:16, 4:5)
- Explainers and course videos
- Kinetic typography and logo stings
- 3D and particle scenes
- AI shots with photoreal characters
- Website and app walkthroughs
- The same video in several languages
Is this for you?
It’s for you if you
- work in marketing, sales or social and need more video
- run a business, agency or product that needs promos
- teach or create content and want better visuals
- already use Claude Code, Codex or Cursor, or want to start
It’s not for you if you
- want a button that makes a talking avatar for you
- don’t plan to try anything yourself after the workshop
YOUR INSTRUCTOR
I’m Davor Debrecin.
I run Digitalni Superheroj and AI Akademija in Croatia, where I teach business owners and professionals to work with AI agents. I made every video on this page with this workflow, and in the workshop I show you exactly how.
I show what works, where I got stuck and how I fixed it, so you don’t have to repeat my mistakes.
What my students say (translated from Croatian)
“We work on concrete examples right away. I especially value how closely Davor follows AI technology.”
“I applied what I learned at Qualis right away. I’ve followed Davor’s courses since 2018.”
“He explains things so I can actually understand them. Simple, patient and clear.”
Everything you need for your first video
3-hour live workshop
Friday, October 2, 11:00–14:00 New York / 17:00–20:00 Berlin on Zoom. Live Q&A.
Lifetime replay
Watch again whenever you need, at your own pace.
4 HyperFrames templates
Source code of the promo and the three films on this page. Swap the text and images, get your own video.
Prompt library
The prompts I used for scenes, images, video shots and voices.
Model and voice cheat sheet
Which model for which job, what it costs, and which voices sound natural in English, Japanese and more.
Prep guide
Four steps before Friday so we can start making video right away.
WHAT THE MODELS COST
Straight talk about costs
Besides the workshop you only pay for what you use on fal, plus an ElevenLabs subscription for voice and music. Real numbers from our videos:
| 61 s promo: 1 Nano Banana Pro image + 6 s of Veo 3.1 video | ≈ $2.55 |
| “Spark” film (46 s): 7 photoreal shots | ≈ $1.05 |
| “Wish List” film (46 s): 13 clay shots | ≈ $1.95 |
| Voice and music | ElevenLabs plan |
Based on fal’s price list on September 29, 2026: Nano Banana Pro $0.15 per image, Veo 3.1 $0.40 per second.
PRICE
Enroll in Video AI Accelerator
Early-bird until Thursday 23:59 Berlin time$149then $199
Price until the workshop starts$199
Enrollment is closed
- 3-hour live workshop, Friday October 2, 11:00 New York
- Lifetime replay
- 4 HyperFrames templates with source code
- Prompt library and model and voice cheat sheet
- Prep guide and private page
- 14-day money-back guarantee
Price in USD, taxes included. Card payment via Stripe.
Enrollment closes Friday at 11:00 New York time, when we go live. After that the workshop can’t be purchased.
14-day money-back guarantee
If after the workshop you decide it wasn’t useful, email me within 14 days and I refund you in full. No questions.
Questions
Do I need to know how to code?
No. The agent writes the code. You describe what you want and review the result. I show you how to ask for changes in plain sentences.
Which tool do I need?
One of three: Claude Code, Codex or Cursor. If you have none, the prep guide tells you which to pick and what it costs.
How much will I spend on models?
A few dollars on fal for one promo. Exact numbers from our videos are above. Voice and music run on an ElevenLabs plan.
Does it work on Windows?
Yes. HyperFrames runs on Mac, Windows and Linux. You need Node.js 22 and FFmpeg.
What if I can’t attend live?
The replay is yours forever and arrives after the workshop. Enrollment still closes when we go live.
Can I make videos in other languages?
Yes. I show how to take the same video into another language with a new voice and on-screen text.
Can I use the videos commercially?
The videos you make are yours. HyperFrames is open source (Apache 2.0). For images, video and voice, fal’s and ElevenLabs’ terms for your account apply.
How do I get a refund?
Email info@digitalnisuperheroj.com within 14 days after the workshop. We refund the full amount.
By Friday evening you have your first video.
Three hours, one agent and an idea you already have. The early-bird price ends Thursday at 23:59 Berlin time, and enrollment closes when we go live.
Save your seat · $149P.S. Everything in this workflow is days or weeks old. A year from now everyone will make video this way. You can start on Friday.