Solve Capture the Flag challenges by automatically decoding mystery strings to find hidden flags.
Decode multi-layer encoded data where text was scrambled multiple times in sequence.
Quickly identify and reverse common encodings like Base64, Rot13, and hex without manual trial-and-error.
Analyze suspicious or obfuscated text in security research and penetration testing.
Requires Rust compilation, BERT model download, and GPU/CUDA setup for inference performance.
Ciphey is an automatic decoding and decryption tool, you hand it scrambled, encoded, or encrypted text and it figures out what was done to the text and reverses it, all without you needing to know the cipher or encoding method in advance. This is particularly useful in Capture the Flag (CTF) competitions, where participants are often given mysterious encoded strings and must decode them to find hidden flags. The tool works by using a search algorithm (A* search) that tries different decoding methods in a smart order rather than brute-forcing everything. It runs fast because it can quickly identify which encoding is most likely (Base64, Rot13, hex, and so on) using a cipher identification component, and it uses multithreading to run multiple attempts in parallel. It supports 16 decoders as of the README, with more being added. A notable feature is multi-level decoding: if text was encoded multiple times in sequence (for example, Base64 applied, then Rot13 applied on top of that), Ciphey can peel away each layer automatically. It also includes a timer so it does not run forever if the text cannot be decoded. For detecting whether the decoded output is real readable text (called plaintext), it uses a combination of English language statistical checks and an optional AI model based on BERT (a type of natural language model) for higher accuracy. Ciphey is written in Rust and can be installed with cargo install ciphey. It also runs as a Discord bot and is available as a library you can build on. The tool is aimed at security researchers, CTF players, and anyone who regularly deals with encoded data.
Generated 2026-05-18 · Model: sonnet-4-6 · Verify against the repo before relying on details.