Framecast AI is a complete AI headshot SaaS. It trains custom AI models, generates studio quality headshots in minutes, and lets users fine tune every result with a built in magic editor powered by Google Gemini. This is a tour of what actually ships in the codebase.
The core product loop
The product is built around one flow that works end to end:
- A user uploads a few photos of themselves.
- A custom model is trained on those photos.
- The user generates headshots in a chosen style.
- The magic editor refines the result in one click.
- The user downloads the final images.
That loop is what makes the product sellable as a service. Everything else supports it.
What is included
The codebase ships production ready features, not placeholders:
- AI powered headshot generation
- A one click magic editor built on Google Gemini
- A styles pack gallery for different looks
- Multi language support with i18n
- A built in admin analytics dashboard
- Multiple payment gateways out of the box
- Custom branding and white labeling
The stack is Next.js, TypeScript, Tailwind CSS, Supabase, OpenAI, Google Gemini, and Stripe. It is architected so you can deploy it, rebrand it, and start selling headshots without rebuilding the plumbing.
Going from purchase to launch
Clone the repository, connect your Supabase and Stripe keys, set your model credentials, and deploy. The live demo shows the product running, and the product page lists every included capability.
The reason to start from a codebase like this is speed. The hard parts, training the model, generating results, handling payment, are already wired. You focus on brand, pricing, and go to market.
