About LeafCode
Your Code Is Melting Glaciers.
We're Here To Fix That.
Every inefficient loop, every wasteful recursive call, every unoptimized data structure — they all consume real energy in real data centers powered by real coal plants. LeafCode makes the invisible visible: your code has a carbon footprint, and it's bigger than you think.
Built by
A team of 5 students passionate about sustainable software
Make Green Code The Default
We're building a generation of developers who instinctively optimize for sustainability — not as an afterthought, but as foundational practice.
Scale Through Education
When millions of developers default to energy-efficient algorithms, we reduce global data center emissions and make computing accessible where energy is scarce.
Gamify The Future
By turning energy efficiency into competition, we make sustainable coding engaging, measurable, and rewarding for the next generation.
The Problem
Software is responsible for a growing share of global energy consumption. Data centers consume approximately 1-2% of global electricity, and that number is projected to triple by 2030. Yet most developers have no idea how much energy their code consumes.
Traditional coding education focuses on correctness and speed, but ignores efficiency. Students learn to solve problems without understanding that a recursive Fibonacci implementation can consume 5-10× more energy than an iterative one.
The result? Millions of developers writing energy-wasteful code that gets executed billions of times daily across the internet, compounding into gigatons of unnecessary carbon emissions.
How LeafCode Works
Real Energy Measurement
Using machine learning models trained on real hardware data, we estimate the actual energy consumption of your code as it runs. No guesswork — real joules, real watts, real impact.
Competitive Learning
Solve the same challenges as LeetCode, but compete on energy efficiency instead of just speed. Climb the leaderboard by writing smarter, not just faster, code.
Multi-Language Support
Practice in Python, JavaScript, C, or C++. Compare how different languages and approaches affect energy consumption for the same algorithm.
Instant Feedback
See your code's energy consumption in real-time. Learn which patterns waste power and which ones scale efficiently.
Energy difference between optimized vs. naive algorithms
Programming languages supported for diverse learning
Energy measurements using ML-based power estimation
Our Vision
By 2030, we envision a world where energy-aware coding isn't a specialty — it's baseline literacy. Where every CS graduate understands that their algorithms have environmental impact. Where "Does it work?" becomes "Does it work sustainably?"
We're not asking developers to sacrifice performance. We're showing them how marginal gains in efficiency create exponential environmental returns when deployed at internet scale. A 10% energy reduction seems trivial — until you multiply it by a billion users and a trillion compute cycles.
Small optimizations. Exponential impact. That's the LeafCode way.
Built at TU Delft
A Sustainable Software Engineering Project
LeafCode was created as part of the CS4575 Sustainable Software Engineering course at TU Delft, combining machine learning, cloud infrastructure, and gamification to make energy-efficient coding accessible to everyone.