ChatGBeanT

Working demo at https://chatgbeant.com

A multi-model AI chat application with image and video generation, document management, and a rich conversational interface.

Chat with Claude, GPT, Gemini, DeepSeek, and more — all in one place. Generate images and videos from text prompts, organize conversations into groups, and build a personal media library.

Features

• Multi-model chat — Switch between 1000+ AI models from providers like Anthropic, OpenAI, Google, and DeepSeek. Responses stream in real-time.
• Image generation — Generate images from text prompts using FalAI and OpenRouter models. Attach images for image-to-image editing.
• Video generation — Generate videos from text or image prompts with configurable duration, aspect ratio, and quality.
• Documents & media library — Create rich text documents, save generated images and videos, and attach any of them to future conversations for context.
• Thread organization — Pin, group, rename, and search threads. Automatic time-based grouping (Today, Last 7 Days, etc.) plus custom named groups.
• Thread & document context — Attach entire previous conversations or documents to new messages to give the AI additional context.
• BYOK (Bring Your Own Key) — Add your own OpenRouter and fal.ai API keys to unlock pro-tier models on the free plan.
• Admin dashboard — Manage users, tiers, roles, and models. Sync models from OpenRouter and FalAI. Toggle featured and active status.
• Theming — Light, Dark, and System themes.

Tech Stack

Layer	Technology
Frontend	Next.js (App Router), React 19, Tailwind CSS v4
Backend	Convex (database, server functions, file storage)
Auth	Clerk
AI Chat	Convex Agent, OpenRouter
Media Gen	fal.ai, OpenRouter
Monorepo	Turborepo

Project Structure

ChatGBeanT/
  apps/
    web/          — Next.js frontend
  packages/
    backend/      — Convex backend (schema, server functions, actions)
    ui/           — Shared UI components
  docs/           — Documentation

Getting Started

Prerequisites

• Node.js 18+
• A Convex account (free)
• A Clerk account (free)

1. Install dependencies

pnpm install

2. Configure Convex

pnpm setup --workspace packages/backend

This will log you into Convex and prompt you to create a project. Then set the following environment variables in the Convex dashboard:

• CLERK_ISSUER_URL — from your Clerk JWT template (use the "convex" template)
• OPENROUTER_API_KEY — from OpenRouter
• FAL_KEY — from fal.ai (for image/video generation)

3. Configure the web app

Create .env using .example.env as a template:

4. Run the app

pnpm dev

Deploying

Deploy using vercel and convex integration

Documentation

• Functional Requirements — Complete specification of all user-facing features