This is an educational reference, not a substitute for a cryptographic review. Use it to shortlist and understand algorithms — then validate any production decision against current standards, your own threat model, and, for high-stakes systems, a qualified cryptographer.

What it is

crypto::compare is a curated, side-by-side reference for 97 cryptographic algorithms across 17 categories — symmetric encryption, KEMs, signatures, hashes, KDFs, MACs, threshold and multi-party schemes, and more. Every entry carries security estimates, recommendation level, known attacks, deployment notes, and primary-source citations. It pairs with companion tools: an algorithm advisor, safe defaults, migration guides, implementation maps, and interactive labs.

Who it's for

Engineers, architects, and students who need to choose cryptography — not implement primitives from scratch. If you are deciding “which KEM,” “is PBKDF2 still acceptable,” or “what does a post-quantum migration look like,” this is built for you.

Threat model & scope

Recommendations assume a standard adversary: a well-resourced attacker bounded by publicly known cryptanalysis, including a future large-scale quantum computer for post-quantum labels. They do not model nation-state implementation backdoors, supply-chain compromise, or physical coercion.
Security estimates reflect known attacks and public literature at review time. They are guidance, not certification, and can change as cryptanalysis advances.
Scope is algorithm selection and tradeoffs — not protocol design, key management operations, or constant-time implementation, each of which can undo a sound algorithm choice.

How the data is sourced & verified

Every algorithm cites primary sources — NIST FIPS/SP, IETF RFCs, ISO, CRYPTREC, KPQC, GB/T, GOST, DSTU, and peer-reviewed Eurocrypt/CRYPTO proceedings. A build-time check rejects entries lacking primary-source citations, and a separate job keeps the linked interactive labs in sync with the live catalog. The dataset and its full provenance are open source.

Update cadence & freshness

Each algorithm's lastReviewed date records when its claims and citations were last re-verified against primary sources. On June 2026 all 97 entries were re-verified in a single pass (dataset v1.1.0), so they share that date — a genuine source-verification stamp, not a claim that each was independently re-derived on a different day. That pass confirmed primary-source citations resolve and support each claim, refreshed post-quantum standardization status, and corrected stale or broken references.

What that review is and is not: it is automated primary-source re-verification and a cryptanalysis scan — not an expert cryptographer's sign-off. Automated freshness checks flag entries older than 180 days so reviews don't silently lapse. Treat research-grade entries as a prompt to verify more aggressively against your own threat model.

What it is not

Not a crypto library — do not copy parameters into production without verifying against current standards.
Not a compliance attestation or a security audit of your system.
Not exhaustive — niche, deprecated, or regional schemes may be missing or summarized.

Source & contributions

The project is open source. Corrections, citations, and new algorithms are welcome via pull request.

Source & dataset on GitHub ↗ · Security policy ↗ · Interactive lab catalog ↗