Image and video decision

Hong Xuan Kow

A skilled developer can build a narrow AI-driven video generator (prompt→render) in ~30 hours using third-party APIs and FFmpeg, but reproducing Koe's full product (managed premium models, credits marketplace, priority render infra, and enterprise features) is substantially more work.

Visit website
Subscription$12/month ✓ verified
Initial build30 hours
Monthly upkeep8 hours + $100
Evidence2/3 runs agree

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

  • Provide a script or prompt → generate storyboard/shot list → synthesize clips (video frames, TTS/voice) via external generative APIs → assemble and render timeline with FFmpeg → produce captions and upscaled export.

What it still won’t have

  • Proprietary premium models and premium-model access
  • Built-in credit packs and integrated credit marketplace
  • Priority render queue and guaranteed fast rendering
  • Custom LoRA character training and enterprise features (SSO, SLA, dedicated support)
  • Seamless 4K pro-quality rendering with managed infra and scaling

What remains hard

  • Product polish and ongoing maintenance
Read the build prompt

First-year cost

Keep paying

Paying ischeaper in year one.

On cash alone, building overtakes the subscription at 9 seats.

Paid seatsseats

Money you would actually spend

Keep paying

Subscription price × seats × 12

Build it

AI build APIs + hosting

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

Not run yet
Build a minimal AI video generation web app using Node.js + Express, React frontend, Postgres for projects, and FFmpeg for assembly. In scope: accept a text script or prompt, parse it into shots, call external generative APIs for frames/images and TTS, assemble timeline with FFmpeg, generate sidecar subtitles, support 720p–1080p export, user signup, and per-user credit tracking. Out of scope: custom model training, multi-tenant priority queues, enterprise SSO, and a marketplace. Include error handling for API failures, retry/backoff, input validation, and unit/integration tests for API handlers and render pipeline.
How we checked4 sources · 2/3 runs agreed · evidence score 63

How the score was reached

  • Partly verdict base52
  • An open-source build was found+5
  • 4 cited sources+3
  • Price verified on pricing page+3
  • Evidence score63

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 · 4

Every page the run actually retrieved.

Integrity checks

What held up, and what did not.

✓ Price read off the page! 2 of 3 runs agreed; the verdict is the majority✓ Citations limited to fetched pages! 1 moat recorded