One theme. One short ready to post.
Zero subscription.
Last Video Labs writes the script, records the voice-over, picks footage visually verified by AI and edits your video — on your PC, with your own API keys.
The whole pipeline, in one interface
From raw theme to final video — with a human check at every step.
AI looks at every clip before keeping it
For every scene, it checks the thumbnails of candidate videos and discards off-topic ones. Nothing fits? It rephrases the search and checks again. No more random footage that has nothing to do with your text.
✕
✕
✕
✓✍️ The script stays in your hands
Edit it, ask for a targeted rewrite, regenerate. Nothing goes to render without your approval.
🎬 Timeline before rendering
Every scene with its real thumbnail and the voice waveform. Force your own video into any scene.
🎙️ Choice of voices, adjustable speed
Free Microsoft voices or premium ElevenLabs. Language/gender filters, audio preview.
💬 Styled captions
6 styles × 8 fonts, synced to the millisecond with the voice.
🎨 Filters & effects
9 color filters, Ken Burns, vignette, 11 transitions. Rendered locally with ffmpeg.
An interface that guides you
Step-by-step onboarding, real-time cost counter, a preview before every spend.
4 steps, ~2 minutes
Plus a real-time cost counter so you always know exactly what you're spending.
Describe your short
Theme, tone, visual world. Add your own images or videos if you want.
Approve the script
Read it, correct it, steer it. ≈ $0.01 in tokens.
Adjust the scenes
Visual timeline, AI-verified footage, manual tweaks if needed.
Final render
Local ffmpeg editing, 1 to 2 minutes. No watermark, it's yours.
Why not a subscription like everyone else?
Classic shorts generators make their margin on AI credits. We just sell you the tool — period.
Typical SaaS tools
- $20 to $40 / month, forever
- Limited credits, watermark on free tier
- Your media uploaded to their servers
- Token prices marked up 10×
- Service shuts down = tool is dead
Last Video Labs
- One-time purchase, lifetime license
- Unlimited renders, never watermarked
- 100% local — nothing leaves your PC
- Your API keys at real cost (~$0.02/video)
- The tool is yours, for good
One price. One time.
- Full Windows application
- Unlimited videos and renders, no watermark
- AI visual verification of footage
- All voices, styles, filters and formats
- Updates included
Frequently asked questions
Is it really 100% local?
Yes. The app runs on your PC. Your API keys are stored in a local file and only sent to their own provider (Anthropic for the script, Pexels/Pixabay for videos, ElevenLabs if you enable it). Your personal media never leaves your machine.
Why do I need API keys? What does the AI actually do?
That's the "Bring Your Own Key" principle: you pay the AI at the provider's real cost (~$0.02 per video) instead of a subscription that marks up the same tokens tenfold. Pexels and Pixabay keys are free; with Anthropic you fund your own account (~$5 covers hundreds of videos). The app shows a real-time cost counter.
Important: the AI (Claude) is used to write the story, split it into scenes, choose and visually verify the footage, and place your media. It does not generate images or videos: the footage comes from real, royalty-free stock banks (Pexels, Pixabay) — which is exactly what keeps the cost down to a few cents.
Will the AI generate the videos itself someday?
It's planned. The "Bring Your Own Key" architecture makes it easy to add generative video providers (Kling, Veo…) exactly like today's stock banks: you'll plug in your key and be able to generate a custom clip for a given scene. Note: generative video is much more expensive than stock footage (from a few cents to several dollars per clip depending on the provider) — so it will arrive as a per-scene option, not a replacement.
Do I own the videos? Can I monetize them?
Yes. The stories are 100% original (generated, never copied), the footage comes from royalty-free banks cleared for commercial use (Pexels/Pixabay licenses), and the voice is generated. No watermark, no restrictions.
Does it work in English?
Yes — English scripts, English voices (free or premium), synced captions. French and other languages supported by the Microsoft/ElevenLabs voices also work.
What if I want to use my own footage?
You can add your images and clips right from the first step: the AI reserves the right scene for them in the edit, or you can assign them yourself to a specific scene from the timeline.