Transform your sketches into stunning AI-generated images and bring them to life with cinematic animations—all on an infinite canvas.
Built for Cursor Hackathon Singapore 2025
DoodleMorph is an AI-powered creative studio built on top of TLDraw's infinite canvas. Draw a simple sketch, and watch as AI transforms it into professional artwork with just a few clicks. Then, select your generated images to create smooth, cinematic animations using cutting-edge video generation models.
- 🎨 Sketch-to-Image Generation: Draw anything and instantly transform it into high-quality images with custom style prompts
- 🎬 Image-to-Video Animation: Select multiple generated images and create smooth animations with Veo 3.1
- 🖼️ Quick Style Presets: One-click styles like Cartoon, Pixel Art, 3D Render, and Sticker
- 🎯 Contextual Tools: Smart toolbar that adapts based on what you select on the canvas
- 📥 Export Everything: Download your generated images and animations as high-quality files
This project was built for the Cursor Hackathon Singapore 2025 and is powered by incredible AI technologies from our sponsors:
- Fal.AI - Lightning-fast AI model hosting and inference for image generation
- Gemini API - Gemini 2.5 Flash for intelligent prompt enhancement
- Veo 3.1 - Google DeepMind's state-of-the-art video generation model
- Groq - Ultra-fast AI inference with GPT OSS for additional prompt generation
- Next.js 14 - React framework with App Router and API routes
- TLDraw v3 - Infinite canvas SDK powering the drawing interface
- React 18 - UI component library
- TypeScript 5 - Type-safe development
- Fal.AI Client - AI model inference and hosting
- Flux-Dev: High-quality image generation
- Kling Video: Advanced video synthesis
- Google Gemini API - Gemini 2.5 Flash for prompt enhancement and refinement
- Groq AI SDK - Fast LLM inference with GPT OSS
- Vercel AI SDK - Unified AI SDK for model integration
- Veo 3.1 - Cinematic video generation from images
- Draw - Use TLDraw's drawing tools to sketch a character, object, or scene
- Select - Click or drag to select your drawing
- Style - A toolbar appears at the bottom. Choose a quick preset (🎨 Cartoon, 🎮 Pixel Art, 🌟 3D Render) or type a custom prompt (max 50 characters)
- Generate - Click "✨ Generate Image" and wait ~10 seconds
- Enjoy - Your AI-generated image appears on the canvas!
- Generate or Import - Create styled images using the process above, or use existing images
- Select Multiple - Select 2 or more images on the canvas (hold Shift to multi-select)
- Animate - The animation toolbar appears. Click "🎬 Create Animation with Veo"
- Wait - Video generation takes 1-2 minutes for cinematic quality
- Download - Preview your animation and click "📥 Download Animation" to save as MP4
Pro Tip: The order you select images determines the animation sequence!
Live site: https://doodlemorph.vercel.app/ YouTube demo video: https://youtu.be/BiEwaw66fSg
Built with ❤️ by:
- Xingxiang
- Akilesh
- Aldric
Special thanks to:
- Cursor Hackathon Singapore 2025 organizers for this incredible event
- Fal.AI, Google (Gemini & Veo), and Groq for providing cutting-edge AI technologies
- TLDraw team for the amazing canvas SDK
- Vercel for seamless deployment and hosting
- The open-source community for inspiration and tools
Made with 🎨 at Cursor Hackathon Singapore 2025