Inspiration
Images can tell a story, but words can make it immortal. We wanted to create a tool that turns any image into a bilingual poem that captures its emotions, colors, and mood, blending visual art with literature. With Gemini 3’s multimodal capabilities, we saw an opportunity to make AI-generated poetry accessible, creative, and shareable.
What it does
PoetryLens lets users upload an image, and instantly generates a poem in both English and Urdu. The output is displayed in a beautifully styled card with gradients, proper typography, and smooth animations. Users can read, enjoy, and even share the poetry, turning visuals into culturally rich literary experiences.
How we built it
Frontend: React.js with Tailwind CSS and Framer Motion for animations
Backend: Node.js / Next.js API route
AI Integration: Gemini 3 Generative Language API for image analysis → poetry generation
Process:
User uploads an image
Image is converted to base64 on the backend
Request is sent to Gemini 3 with a creative prompt
Gemini 3 returns HTML with English & Urdu poetry
Frontend displays it in a stylish, animated card
Challenges we ran into
Handling multimodal input (images → text) with Gemini 3
Generating accurate Urdu poetry with proper script and line breaks
Creating a UI that looks elegant while supporting bilingual text and animations
Ensuring fast API response for a smooth user experience
Accomplishments that we're proud of
Successfully integrated Gemini 3 multimodal capabilities
Generated culturally rich, emotionally evocative poetry in two languages
Built a responsive and visually appealing interface
Maintained proper Urdu typography and right-to-left alignment
What we learned
How to structure prompts for creative bilingual AI outputs
Handling image upload, base64 conversion, and API calls efficiently
Styling dynamic AI-generated content in a React application
Importance of user experience in AI apps—animations, readability, and presentation matter
What's next for PoetryLens
Add voice narration of the poems in both languages
Enable social sharing with preformatted images of poems
Allow users to customize styles and fonts
Introduce themes or moods to guide Gemini 3 in poem generation
Built With
- api
- express.js
- node.js
- react
- tailwind
Log in or sign up for Devpost to join the conversation.