Baim

A desktop app for generating product catalog images using AI.

Batch-generate image variants from a source photo and text prompts. Pick a model provider (Gemini), configure your API key, upload a product shot, write prompts — Baim handles the rest asynchronously.

Features

• Upload & manage source images in a local library
• Batch generation — one source image × many prompts, each becomes its own job

Quick start

npm install
npx tauri dev

Set a Gemini API key in Settings → API Keys, upload an image, write a prompt, and generate.

Development

Command	What it does
npm run dev	Vite dev server (port 5173)
npm run build	Typecheck + Vite build
npm run lint	oxlint
npx tauri dev	Full desktop app
npx tauri build	Production bundle
cargo check -p baim	Check shared crate only

Prerequisites

• Rust (edition 2021)
• Node.js 20+
• Tauri v2 system dependencies
• Gemini API key from Google AI Studio

Architecture

baim/
├── src/               React 19 + react-router v8 frontend
├── baim/              Shared Rust crate (ImageProvider trait + GoogleProvider)
└── src-tauri/         Tauri v2 backend (Rust, SQLite)

The frontend communicates with the Rust backend via Tauri invoke(). Image generation goes through the ImageProvider trait — providers register themselves in all_providers() and are driven from the settings UI. The active provider and API keys are stored in SQLite; the frontend never sees the key value.