NIST Post-Quantum Cryptography Standardization

Round one

Under consideration were:
(strikethrough means it had been withdrawn)

Round one submissions published attacks

• Guess Again by Lorenz Panny
• RVB by Lorenz Panny
• RaCoSS by Daniel J. Bernstein, Andreas Hülsing, Tanja Lange and Lorenz Panny
• HK17 by Daniel J. Bernstein and Tanja Lange
• SRTPI by Bo-Yin Yang
• WalnutDSA
  • by Ward Beullens and Simon R. Blackburn
  • by Matvei Kotov, Anton Menshov and Alexander Ushakov
• DRS by Yang Yu and Léo Ducas
• DAGS by Elise Barelli and Alain Couvreur
• Edon-K by Matthieu Lequesne and Jean-Pierre Tillich
• RLCE by Alain Couvreur, Matthieu Lequesne, and Jean-Pierre Tillich
• Hila5 by Daniel J. Bernstein, Leon Groot Bruinderink, Tanja Lange and Lorenz Panny
• Giophantus by Ward Beullens, Wouter Castryck and Frederik Vercauteren
• RankSign by Thomas Debris-Alazard and Jean-Pierre Tillich
• McNie by Philippe Gaborit; Terry Shue Chien Lau and Chik How Tan

Round two

Candidates moving on to the second round were announced on January 30, 2019. They are:

Round three

On July 22, 2020, NIST announced seven finalists ("first track"), as well as eight alternate algorithms ("second track"). The first track contains the algorithms which appear to have the most promise, and will be considered for standardization at the end of the third round. Algorithms in the second track could still become part of the standard, after the third round ends. NIST expects some of the alternate candidates to be considered in a fourth round. NIST also suggests it may re-open the signature category for new schemes proposals in the future.

On June 7–9, 2021, NIST conducted the third PQC standardization conference, virtually. The conference included candidates' updates and discussions on implementations, on performances, and on security issues of the candidates. A small amount of focus was spent on intellectual property concerns.

Finalists

Alternate candidates

Intellectual property concerns

After NIST's announcement regarding the finalists and the alternate candidates, various intellectual property concerns were voiced, notably surrounding lattice-based schemes such as Kyber and NewHope. NIST holds signed statements from submitting groups clearing any legal claims, but there is still a concern that third parties could raise claims. NIST claims that they will take such considerations into account while picking the winning algorithms.

Round three submissions published attacks

• Rainbow: by Ward Beullens on a classical computer

Adaptations

During this round, some candidates have shown to be vulnerable to some attack vectors. It forces these candidates to adapt accordingly:

CRYSTAL-Kyber and SABER

may change the nested hashes used in their proposals in order for their security claims to hold.

FALCON

side channel attack using electromagnetic measurements to extract the secret signing keys. A masking may be added in order to resist the attack. This adaptation affects performance and should be considered whilst standardizing.

Selected Algorithms 2022

On July 5, 2022, NIST announced the first group of winners from its six-year competition.

Round four

On July 5, 2022, NIST announced four candidates for PQC Standardization Round 4.

Round four submissions published attacks

• SIKE: by Wouter Castryck and Thomas Decru on a classical computer

Selected Algorithm 2025

On March 11, 2025, NIST announced the selection of a backup algorithm for KEM.

First release

On August 13, 2024, NIST released final versions of its first three Post Quantum Crypto Standards. According to the release announcement:

While there have been no substantive changes made to the standards since the draft versions, NIST has changed the algorithms’ names to specify the versions that appear in the three finalized standards, which are:

• Federal Information Processing Standard (FIPS) 203, intended as the primary standard for general encryption. Among its advantages are comparatively small encryption keys that two parties can exchange easily, as well as its speed of operation. The standard is based on the CRYSTALS-Kyber algorithm, which has been renamed ML-KEM, short for Module-Lattice-Based Key-Encapsulation Mechanism.
• FIPS 204, intended as the primary standard for protecting digital signatures. The standard uses the CRYSTALS-Dilithium algorithm, which has been renamed ML-DSA, short for Module-Lattice-Based Digital Signature Algorithm.
• FIPS 205, also designed for digital signatures. The standard employs the SPHINCS+ algorithm, which has been renamed SLH-DSA, short for Stateless Hash-Based Digital Signature Algorithm. The standard is based on a different math approach than ML-DSA, and it is intended as a backup method in case ML-DSA proves vulnerable.
• Similarly, when the draft FIPS 206 standard built around FALCON is released, the algorithm will be dubbed FN-DSA, short for FFT (fast-Fourier transform) over NTRU-Lattice-Based Digital Signature Algorithm.

On March 11, 2025 NIST released Hamming Quasi-Cyclic (HQC) as the fifth algorithm for post-quantum asymmetric encryption as used for key encapsulation / exchange. The new algorithm is as a backup for ML-KEM, the main algorithm for general encryption. HQC is a code-based scheme using different math than ML-KEM, thus mitigating possible weaknesses should any be found in the lattice-based ML-KEM. The draft standard incorporating the HQC algorithm is expected in early 2026 with the final in 2027.

As of mid-2026, FIPS 206 (FN-DSA, based on FALCON) remained a draft standard. NIST submitted the draft for internal approval on August 28, 2025, and, as of April 2026, the Initial Public Draft had not yet been released pending final internal clearance; the final standard was expected in late 2026 or early 2027. Unlike ML-KEM, ML-DSA, and SLH-DSA, FN-DSA's signing process relies on floating-point Gaussian sampling, which is difficult to implement safely and in constant time; NIST has cited this complexity as the main reason for FN-DSA's longer standardization timeline compared with the other three algorithms.

Additional Digital Signature Schemes

Round One

NIST received 50 submissions and deemed 40 to be complete and proper according to the submission requirements. Under consideration are:
(strikethrough means it has been withdrawn)

Round one submissions published attacks

• 3WISE by Daniel Smith-Tone
• EagleSign by Mehdi Tibouchi
• KAZ-SIGN by Daniel J. Bernstein; Scott Fluhrer
• Xifrat1-Sign.I by Lorenz Panny
• eMLE-Sig 2.0 by Mehdi Tibouchi (implementation by Lorenz Panny)
• HPPC by Ward Beullens; Pierre Briaud, Maxime Bros, and Ray Perlner
• ALTEQ by Markku-Juhani O. Saarinen (implementation only?)
• Biscuit by Charles Bouillaguet
• MEDS by Markku-Juhani O. Saarinen and Ward Beullens (implementation only)
• FuLeeca by Felicitas Hörmann and Wessel van Woerden
• LESS by the LESS team (implementation only)
• DME-Sign by Markku-Juhani O. Saarinen (implementation only?); Pierre Briaud, Maxime Bros, Ray Perlner, and Daniel Smith-Tone
• EHTv3 by Eamonn Postlethwaite and Wessel van Woerden; Keegan Ryan and Adam Suhl
• Enhanced pqsigRM by Thomas Debris-Alazard, Pierre Loisel and Valentin Vasseur; Pierre Briaud, Maxime Bros, Ray Perlner and Daniel Smith-Tone
• HAETAE by Markku-Juhani O. Saarinen (implementation only?)
• HuFu by Markku-Juhani O. Saarinen
• SDitH by Kevin Carrier and Jean-Pierre Tillich; Kevin Carrier, Valérian Hatey, and Jean-Pierre Tillich
• VOX by Hiroki Furue and Yasuhiko Ikematsu
• AIMer by Fukang Liu, Mohammad Mahzoun, Morten Øygarden, Willi Meier
• SNOVA by Yasuhiko Ikematsu and Rika Akiyama
• PROV by Ludovic Perret, and River Moreira Ferreira (implementation only)

Round Two

NIST deemed 14 submissions to pass to the second round.

Round Three

Nine candidates advanced from Round Two to Round Three:

NIST has indicated that at least the multivariate candidates will probably require a further round of evaluation.

Anthropic weakened HAWK with Claude Mythos, so HAWK was withdrawn from the competition, as parameter adjustments would make it incompetitive.