Image and video decision
Adobe Firefly
A capable engineer can reproduce a narrow Firefly-like workflow (prompt → generate → edit → export) using open models and standard cloud infra, but Adobe's integrations, brand, and packaged multi-model catalogue make the full paid product hard to match.
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
- Prompt -> run model(s) -> receive generated media -> basic editing tools -> export/store assets.
What it still won’t have
- Adobe-brand integrations across Creative Cloud apps (Photoshop, Premiere, etc.)
- Polished UX, presets, and product polish at scale
- Branded trust and marketing reach
- Centralized multi-provider model catalogue and licensing management
- Enterprise support and SLAs
What remains hard
- Integration maintenance
All the best models, all in one place.
- Brand trust
"Creative Cloud lets me create effortlessly and allows me to focus on what's important."
First-year cost
Keep paying
Paying is—cheaper in year one.
On cash alone, building overtakes the subscription at 31 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 web app (React frontend, Node/Express backend, Postgres, S3-compatible storage) that provides: 1) a prompt form to submit text and optional image inputs, 2) a job queue that calls an open-source image/audio/video inference endpoint (self-hosted or third-party), 3) storage of generated assets and metadata, 4) simple in-browser editing (crop/trim/re-run), and 5) export/download and basic license tagging. Out of scope: reimplementing model training or matching Adobe UI polish and multi-provider catalogue. Include API error handling, input validation, basic unit tests for backend endpoints, and deployment scripts for a single GPU inference host plus a small cloud VM for the app.
How we checked
How the score was reached
- Partly verdict base52
- An open-source build was found+5
- 4 cited sources+3
- 3/3 assessment runs agreed+4
- Evidence score64
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.
- official productAdobe homepage
- official pricingCreative Cloud plans
- open sourceAnil-matcha/Open-Generative-AI
- open sourceOsmantic/ODS
Integrity checks
What held up, and what did not.
