Image and video decision
Luma Dream Machine
Build a narrower self-hosted workflow for generating and iterating assets using third‑party models and simple agent orchestration, but you cannot fully replace Luma's proprietary models and production-grade capabilities without access to their research-backed models.
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
- Orchestrate third‑party image/video/audio models to generate and iterate assets from prompts, store assets and context, and export deliverables.
What it still won’t have
- Proprietary Luma models (Ray3.2, Uni-1) and research-driven model improvements
- Integrated high-fidelity production video quality and credits pricing tied to Luma models
- Enterprise features (SSO, usage analytics, dedicated fine-tuning) and team billing/credit pooling
- Polished agent orchestration and built-in multimodal consistency across formats
What remains hard
- Proprietary models
We build the models underneath Luma, and frontier research ships straight into the hands of working creatives.
- Proprietary models
UNI-1 Uni-1 is brand intelligence at the model level.
First-year cost
Keep paying
Paying is—cheaper in year one.
On cash alone, building overtakes the subscription at 7 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 lightweight self-hosted creative agent frontend and orchestration service using Next.js (React) + Node/Express backend, Postgres for metadata, MinIO or S3 for assets, and BullMQ workers. Core features in scope: 1) Web UI to create projects, prompts, and templates; 2) Integrations with external generation APIs (image/video/audio + ElevenLabs) with pluggable adapters; 3) Background job queue for long-running renders with status/webhook handling; 4) Credit/usage tracking and simple billing dashboard; 5) Export of generated assets (MP4, JPG/PNG, WAV) and basic versioning. Out of scope: training or reproducing proprietary models (Ray3.2/Uni-1), enterprise SSO, multi-tenant team billing, and advanced production EXR pipelines. Include error handling for failed API calls, retry logic, tests for API integrations and worker flows, and basic CI to run tests.
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
- Hard moats found in the evidence-3
- Evidence score60
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 productLuma | AI Agents for Creative Work
- official pricingPlans & Pricing | Luma
- official productCreative agents that make you prolific | Luma
- open sourceAaronFeng753/Waifu2x-Extension-GUI
- open sourcehacksider/Deep-Live-Cam
Integrity checks
What held up, and what did not.
