Faceless video studio · runs on your PC
Type a topic. Get a finished Short with script, neural voice, AI images, motion, music and word-synced subtitles. It all renders on your own GPU: no subscription, no limits.
A real Short made with the app in 4 minutes. Tap to unmute.
The studio
A local web app that opens in your browser. Every scene is a card you can rewrite, reorder or regenerate on its own.
How it works
Write one line, or paste an article or a story. The script follows hook, turn and payoff, and refuses to invent dates.
Edit a line and only that voice regenerates. Edit a prompt and only that image does.
Neural voices in Spanish and English. SDXL images with a Fast mode (10 s each) and a High mode (80 s) for scenes with people.
Camera motion, crossfades, ducked music, karaoke subtitles, watermark and outro. Thumbnail included. Archive to a drive in one click.
What's inside
Your scripts and videos never leave your computer. No cloud queue, no credits.
Groups respect punctuation. Karaoke highlight on the spoken word. Six free fonts, size, position and colors.
Drop your tracks by mood. Music dips about 10 dB under the voice and comes back in the pauses.
Fast for drafts, High for faces. Regenerate any single scene with a new seed.
Replace any generated image with a photo or a clip. It is cropped, timed to the voice and muted for you.
Translate a finished script in one click, keeping the images. Two channels from one story.
Auto thumbnail with a highlighted keyword, your logo as watermark, and a branded outro.
Every video with its state. Duplicate for variants, trash with undo, archive to an external drive and free the disk.
It is a local web page. Open it from your phone on the same Wi-Fi while the laptop renders.
Made with it
The app was built to run a Spanish-language channel of true stories. Every Short there comes out of it.
The surgeon who operated on 60 people without operating
The telegram where Einstein answered whether he believed in God
The scientist who disappeared after saying reality is an illusion
Requirements
| OS | Windows 10 / 11 (64-bit) |
| GPU | NVIDIA with 6 GB VRAM or more (RTX 3050 laptop tested). 12 GB makes High mode faster. |
| RAM | 16 GB (24 GB recommended) |
| Disk | 20 GB for the models, plus ~200 MB per video until you archive it |
| Internet | Only for the neural voices and to download the models once |
Pricing
Cloud video tools charge $20–60 every month and cap your minutes. Dcrypto Studio runs on hardware you already own, so one license covers unlimited videos. Early buyers get every update to version 1.
Launch price for the first 100 licenses. No spam: one email when it ships.
FAQ
Yes. Images are generated by open-weight models on your GPU, voices use free neural TTS, and rendering is ffmpeg. The only cost is electricity. Scripts can be pasted from any assistant, or generated with your own API key for cents.
Yes. The image model's community license allows commercial use under $1M in yearly revenue, the fonts are OFL, and the scripts are yours. Tick the "altered or synthetic content" box when uploading; platforms reward declaring it and penalize hiding it.
Not if the script is yours. That policy targets mass-produced, templated uploads. Original narration with generated visuals is allowed; the app's script rules push you toward hooks, verified facts and a point of view.
Bring your own tracks. We recommend the YouTube Audio Library, the only source YouTube guarantees won't be claimed. Drop files by mood and the app picks them and ducks them under the voice.
Voice, subtitles and rendering work on any PC. Image generation needs CUDA today; a cloud option for images is on the roadmap. Join the waitlist and tell us your setup.