#8 · Primary category: Cybersecurity & Decryption Tools
magika
Fast and accurate AI powered file content types detection
Project last updated:08/21/26
GitHub Stars
18.0K
Forks
1.1K
Contributors
45
License
Apache-2.0
Why we included this project
Magika tackles a question that quietly breaks a lot of file-handling code: what kind of file is this, actually? Instead of trusting extensions or hand-written magic-byte rules, it runs a small deep-learning model, a few megabytes in size, that identifies content types in milliseconds on a single CPU across more than 200 binary and text formats. That makes it a natural fit for upload handlers, email filters, and malware triage, where a mislabeled file can cause real trouble. It ships as a command-line tool with Python, Node, and Go bindings, so teams can drop it into existing pipelines without changing their architecture. Google already uses it at scale to route files in Gmail, Drive, and Safe Browsing, and it has been integrated into VirusTotal and abuse.ch.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
reverse-skill
AI-powered skill router for reverse engineering, authorized penetration testing, and security research, with on-demand toolchain bootstrapping and self-evolving knowledge base.
gitleaks
Find secrets with Gitleaks 🔑
osquery
SQL powered operating system instrumentation, monitoring, and analytics.
NeoPass
Your Essential Exam Companion for the Iamneo Portal & NPTEL Exams Disguised as NeoExamShield bypass
anubis
Weighs the soul of incoming HTTP requests to stop AI crawlers