Eikon Studio

An open-source AI image generation platform built with Next.js and Convex, with support for multiple image models. Generate images from text, edit existing images, organize them in a personal gallery, and track usage analytics — all in one place.

Live demo: eikonstudio.xyz

---

Features

• Text-to-image and image-editing modes
• Three integrated image variants: Nano Banana 2, Nano Banana Pro, and OpenAI GPT Image 2 — switch per generation
• Public, source-backed /models catalog spanning all ten Eikon model families
• Personal gallery with folder organization
• Usage analytics dashboard
• Google OAuth authentication (via Better Auth + Convex)
• Public REST API for programmatic image generation
• Real-time sync powered by Convex

Tech Stack

• Frontend: Next.js 16 (App Router), React 19, TypeScript, Tailwind CSS v4
• Backend: Convex (real-time database + serverless functions)
• Auth: Better Auth with the @convex-dev/better-auth component (Google OAuth)
• AI: Google Gemini (gemini-3.1-flash-image, gemini-3-pro-image) and OpenAI (gpt-image-2)
• UI: Radix UI + shadcn/ui patterns

Model catalog and execution readiness

The canonical registry in @eikonstudio/core is checked against first-party provider documentation as of 2026-08-02. It records every verified variant, including preview, deprecated, entitlement-restricted, and not-yet-executable models. A provider listing a model does not mean Eikon can execute it.

Only these catalog entries are currently marked Eikon ready and appear in the image creator:

• gemini-3.1-flash-image — Nano Banana 2
• gemini-3-pro-image — Nano Banana Pro
• gpt-image-2 — GPT Image 2

All other entries remain visible at /models as discovered metadata and are excluded from generation selectors until a provider adapter is actually integrated.

---

Getting Started

Prerequisites

• Node.js 20+
• pnpm 10.18.3+
• A Google Gemini API key
• (Optional) An OpenAI API key if you want to use GPT Image 2
• A Convex account (free tier works)
• Google OAuth credentials from the Google Cloud Console

1. Clone and install

git clone https://github.com/MohtashamMurshid/eikonstudio.git
cd eikonstudio
pnpm install

2. Configure environment variables

touch apps/web/.env.local

Fill in the required values in apps/web/.env.local.

3. Set up Convex

In a separate terminal, run:

pnpm dev:convex

This provisions a dev deployment, writes CONVEX_DEPLOYMENT / NEXT_PUBLIC_CONVEX_URL into apps/web/.env.local, and keeps the backend in sync with your apps/web/convex/ folder.

Provider credential encryption: configure a unique 32-byte base64 secret on the Convex deployment before saving or resolving provider credentials:

openssl rand -base64 32
pnpm --dir apps/web exec convex env set CREDENTIAL_ENCRYPTION_SECRET '<generated-value>'

Existing pre-v2 records are read-only compatible only when LEGACY_CREDENTIAL_ENCRYPTION_SECRET is explicitly set to the deployment's former ENCRYPTION_SECRET value. There is no default/fallback secret. Re-save each legacy credential to rewrite it as AES-256-GCM, then remove the legacy secret after inventory confirms no legacy records remain.

4. Configure Google OAuth

In the Google Cloud Console:

1. Create an OAuth 2.0 Client ID (Web application).

2. Add http://localhost:3000/api/auth/callback/google as an authorized redirect URI for local development. Add https://<your-production-domain>/api/auth/callback/google separately for production. This Next.js app proxies auth through /api/auth, so Google returns to the app's origin.

3. Configure GOOGLE_CLIENT_ID and GOOGLE_CLIENT_SECRET on the corresponding Convex deployment. Values in apps/web/.env.local are not automatically copied to Convex.

4. Set the development Convex deployment's auth base URL:

   pnpm --dir apps/web exec convex env set SITE_URL http://localhost:3000

   Keep NEXT_PUBLIC_SITE_URL=http://localhost:3000 in apps/web/.env.local, and keep the production deployment's SITE_URL set to its HTTPS domain. convex/auth.ts reads SITE_URL from Convex; DEV_SITE_URL is not used. If local Google sign-in redirects to production, check this deployment setting first, then begin a fresh sign-in from /auth.

5. Run the app

pnpm dev

Open http://localhost:3000.

---

Scripts

Command	Description
pnpm dev	Start workspace development tasks via Turbo
pnpm dev:convex	Sync the Convex backend in development
pnpm codegen	Regenerate Convex TypeScript bindings
pnpm build	Build all workspace packages via Turbo
pnpm start	Run the web production build
pnpm lint	Lint all workspace packages via Turbo
pnpm test	Run package contract tests via Turbo
pnpm typecheck	Type-check all workspace packages via Turbo

---

Project Structure

apps/
  web/
    app/          Next.js App Router pages and API routes
    convex/       Convex schema, queries, mutations, and auth wiring
    components/   React components (UI, gallery, image combiner, dashboard)
    lib/          Shared utilities (auth, cost calculation, secure storage)
    hooks/        Custom React hooks
    public/       Static assets
packages/
  core/           Shared IDs, schemas, model registry, and job lifecycle
  providers/      Provider-adapter interfaces and contract fixtures
pnpm-workspace.yaml
turbo.json

---

Public API

Eikon Studio exposes a public REST endpoint for programmatic image generation. Each request must include a platform API key generated from your Eikon Studio account (Studio → Settings → API Keys).

Endpoint

POST /api/v1/generate

Authentication

Send your platform API key as a Bearer token or in the x-api-key header:

Authorization: Bearer eik_your_platform_key
# or
x-api-key: eik_your_platform_key

Body parameters

Field	Type	Required	Default	Description
prompt	string	yes	—	Description of the image to generate
provider	string	yes	—	"gemini" or "openai"
model	string	no	provider default	Specific model id to use
imageSize	string	no	"2K"	"1K", "2K", or "4K"
aspectRatio	string	no	"square"	"square", "portrait", "landscape", "wide"

Example

curl -X POST https://eikonstudio.xyz/api/v1/generate \
  -H "Authorization: Bearer eik_your_platform_key" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "A beautiful landscape with mountains and a lake at sunset",
    "provider": "gemini",
    "imageSize": "2K",
    "aspectRatio": "landscape"
  }'

The endpoint is CORS-enabled and can be called from any origin. Requests are authenticated per-user, but there is currently no rate limiting — if you self-host Eikon Studio, you are responsible for your own Gemini / OpenAI usage and costs. Consider adding rate limiting before exposing a public deployment.

---

Deployment

Any platform that runs Next.js works. The live demo is deployed on Vercel:

1. Push the repo to GitHub.
2. Import the project into Vercel with the repository root as the project root.
3. Add the required web environment variables and CONVEX_DEPLOY_KEY to the deployment. The root vercel.json installs the workspace, runs Convex and the Next.js build from apps/web, and publishes apps/web/.next.
4. Configure the production CONVEX_URL / CONVEX_SITE_URL and OAuth callback origins for the deployed domain.

---

Contributing

Contributions are welcome. Please open an issue to discuss major changes before submitting a pull request.

License

MIT © Mohtasham Murshid