Image and video decision
Captions
A small team can reproduce a useful subset (auto-captions, simple AI-guided cuts, export) at modest cost; the full product (generative avatars, managed high-volume credits, polished multi-model stack) is harder to match, so consider building a narrow replacement and keep paying for the premium features you need.
Visit website↗Open-source builds that already do this
Every project below is open source and already does this job today. Fork one, self-host it, or take the parts you need — the build prompt further down assumes an empty file, and this is the shortcut past that. Licences differ; check the one on each card before you ship.
What a replacement has to do
- Upload footage -> auto-transcribe and detect cuts -> apply style-based cuts and captions -> render/export final video
What it still won’t have
- Generative AI avatars / digital twins
- In-app AI video generation (creating new footage)
- Proprietary multi-model editing stacks and high-volume credit tiers
- Eye-contact correction and some advanced denoise/voice-clone features
- 100+ built-in edit styles and managed asset library
What remains hard
- Brand trust
20M current users
First-year cost
Keep paying
Paying is—cheaper in year one.
On cash alone, building overtakes the subscription at 9 seats.
Money you would actually spend
Time you would spend
—
What you would spend
What we assumed
The verdict above measures whether you could build it. This one is only about money.
Runnable build prompt
Build a minimal AI video-editing web app using React (frontend) + Node/Express (backend) + PostgreSQL for metadata and S3 for media. Core features in scope: 1) upload and store video files, 2) transcode and run scene-detection with FFmpeg, 3) generate time-aligned captions via Whisper or AssemblyAI, 4) create an edit plan (simple heuristics + prompt to an LLM) that trims silencers and applies a chosen style, 5) render final MP4 with FFmpeg and provide download. Out of scope: generative video/AI avatars, advanced eye-contact correction, large-scale credit management, mobile apps. Include error handling for failed transcodes and API calls, unit tests for backend endpoints, and an end-to-end test for upload→export flow.
How we checked
How the score was reached
- Partly verdict base52
- An open-source build was found+5
- 5 cited sources+3
- Price verified on pricing page+3
- 3/3 assessment runs agreed+4
- Evidence score67
The base comes from the verdict. Everything under it is a check that either happened or did not, and each one is a fact frozen in this record rather than a judgement made at render time — so the same evidence always produces the same number.
How scoring works →Cited sources · 5
Every page the run actually retrieved.
- official productCaptions: AI That Edits Like a Professional Editor
- official pricingCaptions AI Pricing and Subscription Plans | Captions
- official docsProduct Video Maker: Create Demos That Convert | Captions
- open sourcezhouxiaoka/autoclip
- open sourcepalmier-io/palmier-pro
Integrity checks
What held up, and what did not.
