TrashTalk 2.0 - Compare coding profiles, pull receipts & roast your friends

Your friend solved 500 LeetCode Easy problems and calls himself a 10x engineer. Time to check the receipts. Drop any two LeetCode, Codeforces, GitHub, or CodeChef usernames into TrashTalk 2.0 to see who is actually better. It checks who solves real Hard problems versus who just farms Two Sum, shows the problems you beat them on, roasts the loser based on real stats, and gives you a scoreboard card to drop in the group chat.

Add a comment

Replies

Best

Hey Hey Guys,

Every developer friend group has that one guy who solved 400 LeetCode Easy problems, fixed two typos in a README for green GitHub squares, and talks like he wrote the Linux kernel.

I built TrashTalk 2.0 () to settle those arguments with actual receipts.

You drop in any two usernames from LeetCode, Codeforces, GitHub, or CodeChef, and instead of just comparing one surface-level number, it pulls their real history and exposes the truth:

  • No Easy-Problem Farming: Solving 500 Easy problems will not save you if your friend is actually cracking Hard DP and graph problems or competing in rated contests.

  • No Coasting on Old Glory: If someone hit a high contest rating two years ago and retired, they lose points to someone actively shipping code and solving problems over the last 6 months.

  • The Trophy Room: It finds the exact problems both of you solved, and more importantly, lists the hardest problems you solved that your friend could not figure out.

  • Roasts Backed by Data: Once the scoreboard locks in a winner, it writes a roast for the loser using their actual stat gaps (and there is a solo roaster at /roast if you just want to humble yourself before an interview).

  • Group-Chat Ready Receipts: Generates a clean scoreboard image on the server so you can paste the evidence straight into Discord, Slack, or X before they make their profile private.

The Backstory (2020 vs. Now):

I originally built the first version of TrashTalk back in 2020 in my college dorm using Python Flask and BeautifulSoup web scraping. Every time a coding site changed a single div class, my app crashed. Six years later, I decided to keep the original 2020 code untouched in a /legacy folder as a museum piece and rebuild the whole thing properly in Next.js and TypeScript.

The entire project is free and open source on . Run a matchup against a coworker or friend and let me know who took the L!

How does your product turn technology into positive change? Share the cause you're supporting, the people you're helping, and the impact you hope to make. Tell us what you deployed on Cloud Run and how it works together. Which of these did you use? - Deploy and host a web app - Host LLMs with Cloud Run GPUs - Host AI agents - Build microservices - Data processing
Deploy and host a web app