Overview pranAI is a Chrome extension that helps users understand the environmental impact of their AI prompts. It estimates the energy and carbon cost of each request, suggests more efficient rewrites, and tracks total energy saved with real-world comparisons. Goal: Our goal: make sustainable AI use transparent, measurable, and fun.

Inspiration

As AI becomes part of daily life, few people realize the hidden energy cost behind every prompt. Research shows that running a large language model (LLM) query can use as much energy as charging a smartphone multiple times — and when millions of people use AI daily, that adds up to enormous carbon emissions. We wanted to answer one question: “What if people could see the impact of their AI usage — and actually do something about it?” That idea became pranAI: a lightweight, accessible tool that empowers users to make mindful choices while still enjoying the power of AI. Key Features Prompt Energy Estimation: Calculates approximate energy and CO₂ footprint for each AI request. Smart Rewrite Suggestions: Uses linguistic heuristics to produce a shorter, equally effective prompt. Impact Dashboard: Tracks energy saved over time and displays real-world equivalents Privacy First: All analysis runs locally; no data is collected or stored externally.

How we built it

pranAI is built as a Chrome extension using: Main Languages: HTML, CSS, JavaScript Extension Logic: Chrome Extension API (Manifest v3) Data Sources: Public LLM energy token usage and CO₂ equivalency datasets When a user submits a prompt: The extension captures the input text. We estimate token count and computational complexity. Using benchmark data, we approximate energy use We convert that energy into CO₂ equivalents and relatable comparisons The system then offers an optimized rewrite by simplifying sentence structure and removing redundancies.

Challenges we ran into

Balancing accuracy and usability: We wanted energy estimates that felt meaningful without overwhelming users with scientific detail. Prompt optimization: Generating cleaner prompt suggestions without changing meaning required nuanced linguistic logic. Browser limitations: Handling dynamic text fields in AI chat interfaces (like ChatGPT’s textbox) required creative DOM manipulation.

Accomplishments we're proud of

Built a fully functional Chrome extension in under 48 hours that integrates seamlessly with AI chat platforms. Translated complex energy data into clear, engaging insights through a user-friendly dashboard. Developed a working energy estimation model and prompt optimizer that reduced estimated energy use by up to 30%. Sparked awareness about sustainable AI and how everyday users can make a measurable impact.

What we learned

In this project, we discovered how small text changes can significantly affect AI computation costs while gaining hands on experience with Chrome Extension development. We learned to translate abstract data (like joules and emissions) into relatable remarks people actually care about. Most importantly, we learned that sustainability in AI isn’t just a hardware or data center problem — it’s also a user problem.

What's next

Add gamification — badges, leaderboards, and social challenges for sustainable AI users. Create an pranAI API for developers to integrate sustainability metrics into their own AI tools. Partner with carbon-offset organizations so virtual savings can create real-world environmental impact.