Launched this week
AI Code Checker
productivity
1 follower
productivity
1 follower
AI Code Checker is an online AI code detector for teachers, TAs, interviewers and developers who need to check code for AI signals without treating a single score as proof of authorship. Paste a snippet or upload a file, and this AI-generated code checker returns a 0–100 AI-likelihood score with clear bands: 50 and above is likely AI-generated, 30–49 is worth a manual review, and below 30 shows no clear AI signals.
Launch tags:Productivity
Launch Team
Render Sandboxes: a computer for every agentRun untrusted code safely. Idle sandboxes cost nothing
Promoted
Hey Product Hunt! 👋
I’m the maker behind AI Code Checker (https://aicodechecker.org).
Why I built this
If you teach a CS class or run technical interviews, you’ve probably hit the same frustration: someone turns in super clean, textbook-looking
code, and generic AI text detectors either throw a black-box 94% AI score with zero explanation—or flag well-formatted human code just
because it uses Black/Prettier.
You shouldn’t accuse a student or reject a candidate based on an unexplained percentage. So I built AI Code Checker as a code-review triage &
evidence tool, not a “guilty/innocent” verdict machine.
What makes it different:
• Line-level evidence (not just a score): Every check highlights specific line ranges, exact quoted snippets, and plain-English
reasons—validated against your source lines so the model can't hallucinate line numbers.
• Follow-up questions to verify authorship: Instead of trusting a score blindly, you can generate targeted questions about specific lines in
the submission to ask the student or candidate in a quick 2-minute follow-up.
• Built for real class & hiring workflows: Paste a snippet or upload a ZIP archive (up to 50 files on Pro, 200 on Business). It automatically
ignores lockfiles/node_modules/build artifacts, groups results by student folder, lets you sort by score, and exports CSV or clean PDF
reports.
• Conservative by design: Short snippets (<10 non-empty lines) won't be force-scored, and the prompt explicitly ignores tidy formatting alone.
When evidence isn't clear, it tells you so.
• Fair credit rules: View real human vs. AI sample reports without signing up, get 2 free credits on sign-up (no card required), never get
double-charged for re-checking the exact same code, and automatically get refunded if a check fails.
What’s included:
• Free: 2 credits on sign-up (up to 15,000 chars/check) + full sample reports
• Pro ($9/mo or $90/yr): 300 credits/mo, up to 75,000 chars/check, 50-file ZIP batches, student folder grouping, CSV/PDF export
• Business ($29/mo or $290/yr): 1,200 credits/mo, 200-file ZIP batches, 24-month retention, and REST API (/api/v1) access
I’d love your honest feedback—especially if you’re an instructor, TA, or hiring manager:
1. Do the line-level reasons and follow-up questions match what you’d actually look for when reviewing suspicious code?
2. Which language, IDE/LMS integration, or false-positive edge case should we tackle next?
Drop a sample snippet or test a ZIP and let me know where it breaks! 🙏