Skip to content
ContentLora

    Tip: press / anywhere to search.

    product

    AlphaFold

    Also known as AlphaFold2, AlphaFold 3, AlphaFold Protein Structure Database, AFDB

    AlphaFold is a family of deep-learning models from Google DeepMind that predict the 3D structure of proteins and, since AlphaFold 3, of complexes with DNA, RNA and small molecules.[1][2] Its developers shared the 2024 Nobel Prize in Chemistry, and its open database had more than 3.4 million users by 2026.[3][4]

    Editor reviewedUpdated AI for scienceLife sciencesArtificial intelligence
    Key facts

    AlphaFold, an AI system from Google DeepMind, made accurate protein structure prediction routine.[5] AlphaFold2 matched experimental accuracy for most targets in the CASP14 blind test, and the work earned Demis Hassabis and John Jumper half of the 2024 Nobel Prize in Chemistry.[1][3]

    Versions

    AlphaFold2 (paper published July 2021) combined deep learning with physical and biological knowledge and used multiple sequence alignments of related proteins.[6] It was trained on a 2019 copy of the protein-data-bank.[7]

    AlphaFold 3 (Nature, May 2024) moved to a diffusion-based architecture that predicts whole complexes: proteins with DNA, RNA, small molecules, ions and modified residues.[2] Its paper reported far greater accuracy on protein–ligand interactions than docking tools and better antibody–antigen predictions than AlphaFold-Multimer.[8]

    Impact

    The Nobel committee said AlphaFold2 has been used to predict the structures of virtually all 200 million known proteins and had more than two million users in 190 countries by 2024.[9] Before it, experiments had covered about 100,000 unique proteins.[10] By March 2026 the AlphaFold Database, run with EMBL-EBI, had more than 3.4 million users.[4] In October 2026 the database said it offered open access to over 260 million predictions.[11] It also groups predictions into curated collections, for example for pandemic preparedness, antimicrobial resistance, neglected tropical diseases and the 30 WHO priority global health proteomes.[12]

    2026: from single chains to complexes

    In March 2026 EMBL-EBI, Google DeepMind, NVIDIA and Seoul National University added 1.7 million high-confidence predicted homodimers (two copies of a protein bound together) to the database.[13] The RCSB PDB now lists over a million computed models from AlphaFold DB and ModelArchive next to its experimental entries.[14]

    Access and open alternatives

    AlphaFold 3 is less open than AlphaFold2. Its inference code is on GitHub, but the model parameters may only be used if obtained directly from Google under terms of use. Google also runs a free AlphaFold Server for non-commercial use, with a more limited set of ligands and modifications.[15] Isomorphic Labs, the drug-discovery company Hassabis also leads, says it released AlphaFold 3 together with Google DeepMind.[16][17]

    Academic groups have built open alternatives. The Boltz team describes Boltz-1 as the first fully open model to approach AlphaFold 3 accuracy and releases code and weights under the MIT licence; its successor Boltz-2 adds binding-affinity prediction.[18][19] OpenFold3-preview, from Columbia University and the OpenFold consortium, aims to be a bitwise reproduction of AlphaFold 3 under the Apache 2.0 licence.[20]

    Limits

    Predictions are hypotheses, not measurements. CASP17 in 2026 deliberately targets areas where deep learning has yet to deliver, including immune complexes, ligand binding and multiple conformations.[21] Google DeepMind also stepped in to fund CASP temporarily in 2025 when its NIH grant ran out; STAT reported the one-time gift would cover about 12 months.[22][23] CASP17 lists Google DeepMind as a supporter alongside its NIH sponsor.[24]

    Demis Hassabis, who shared the Nobel Prize for AlphaFold, also leads the drug-discovery company isomorphic-labs.[17] For protein design, see rfdiffusion.

    Questions readers ask

    What problem does AlphaFold solve?

    It predicts a protein's 3D structure from its amino-acid sequence, which the Nobel committee called a 50-year-old problem.[25]

    How accurate is AlphaFold?

    In the CASP14 blind test AlphaFold2 was competitive with experimental structures in a majority of cases.[1]

    What does AlphaFold 3 add?

    It predicts complexes of proteins with nucleic acids, small molecules, ions and modified residues using a diffusion-based architecture.[2]

    Can anyone use AlphaFold 3?

    Its inference code is public, but the model weights may only be used if obtained directly from Google under terms of use. A free AlphaFold Server is available for non-commercial use, with a more limited set of ligands.[15]

    Are there open alternatives to AlphaFold 3?

    Yes. Boltz-1 is described by its developers as the first fully open model to approach AlphaFold 3 accuracy, and OpenFold3-preview aims to reproduce AlphaFold 3 under the Apache 2.0 licence.[18][20]

    What changed in the AlphaFold Database in 2026?

    In March 2026 EMBL-EBI and partners added 1.7 million high-confidence predicted homodimer structures.[13]

    Sources

    Each numbered claim is a statement we checked against the sources listed with it. Status shows how well established it is.

    1. [1]

      AlphaFold2, published in Nature in July 2021, was validated in the CASP14 blind assessment and predicted structures with accuracy competitive with experiment in a majority of cases. confirmedas of 2021-07-15

    2. [2]

      AlphaFold 3, published in Nature in May 2024, uses a diffusion-based architecture to predict joint structures of complexes containing proteins, nucleic acids, small molecules, ions and modified residues. confirmedas of 2024-05-08

    3. [3]

      The 2024 Nobel Prize in Chemistry went half to David Baker for computational protein design and half jointly to Demis Hassabis and John Jumper for protein structure prediction. confirmedas of 2024-10-09

    4. [4]

      As of March 2026 the AlphaFold Database had more than 3.4 million users in 190 countries. confirmedas of 2026-03-16

    5. [5]

      AlphaFold is an AI system developed by Google DeepMind that has predicted the structure of millions of proteins since 2021. confirmedas of 2026-03-16

    6. [6]

      AlphaFold2 combines a deep-learning architecture with physical and biological knowledge about protein structure and uses multiple sequence alignments of related proteins. confirmedas of 2021-07-15

    7. [7]

      AlphaFold2 was trained on a copy of the Protein Data Bank downloaded in August 2019. confirmedas of 2021-07-15

    8. [8]

      The AlphaFold 3 paper reported far greater accuracy for protein–ligand interactions than state-of-the-art docking tools and higher antibody–antigen accuracy than AlphaFold-Multimer. confirmedas of 2024-05-08

    9. [9]

      According to the Nobel committee, AlphaFold2 has been used to predict the structure of virtually all 200 million known proteins and has been used by more than two million people in 190 countries. confirmedas of 2024-10-09

    10. [10]

      Before AlphaFold2, experiments had determined structures for around 100,000 unique proteins, a small fraction of the billions of known protein sequences, because each structure takes months to years of work. confirmedas of 2021-07-15

    11. [11]

      As of October 2026 the AlphaFold Protein Structure Database, developed by Google DeepMind and EMBL-EBI, said it provides open access to over 260 million protein structure predictions. confirmedas of 2026-10-10

    12. [12]

      The AlphaFold Database offers curated collections for pandemic preparedness, antimicrobial resistance, neglected tropical diseases and the 30 WHO priority global health proteomes. confirmedas of 2026-10-10

    13. [13]

      In March 2026 EMBL-EBI, Google DeepMind, NVIDIA and Seoul National University added 1.7 million high-confidence predicted homodimer structures to the AlphaFold Database. confirmedas of 2026-03-16

    14. [14]

      As of 10 October 2026 the RCSB PDB listed 260,626 experimental structures and 1,062,058 computed structure models from AlphaFold DB and ModelArchive. confirmedas of 2026-10-10

    15. [15]

      AlphaFold 3's inference code is public on GitHub, but its model parameters may only be used if obtained directly from Google under terms of use; the model is also offered on AlphaFold Server for non-commercial use, with a more limited set of ligands. confirmedas of 2026-10-10

    16. [16]

      Isomorphic Labs says it released AlphaFold 3 in 2024 together with Google DeepMind. confirmedas of 2026-02-10

    17. [17]

      Demis Hassabis is chief executive of Isomorphic Labs and Max Jaderberg is its president, according to the company's May 2026 funding announcement. confirmedas of 2026-05-12

    18. [18]

      The Boltz team describes Boltz-1 as the first fully open-source model to approach AlphaFold 3 accuracy, and releases Boltz code and weights under the MIT licence for academic and commercial use. confirmedas of 2026-10-10

    19. [19]

      Boltz-2, described in a June 2025 preprint by MIT and Recursion researchers, predicts both complex structures and small molecule–protein binding affinity, and its authors say it is the first AI model to approach free-energy perturbation accuracy while being at least 1,000 times more computationally efficient. confirmedas of 2025-06-18

    20. [20]

      OpenFold3-preview, from Columbia University's AlQuraishi Lab and the OpenFold consortium, aims to be a bitwise reproduction of AlphaFold 3 and is available under the Apache 2.0 licence for academic and commercial use. confirmedas of 2026-10-10

    21. [21]

      CASP17 emphasises areas where deep learning has yet to deliver, including immune complexes, protein–ligand complexes, nucleic acids and conformational ensembles. confirmedas of 2026-10-10

      • CASP17 home page · Protein Structure Prediction Center (UC Davis) (retrieved 2026-10-10)
    22. [22]

      In July 2025 Google DeepMind provided temporary funding to CASP after its NIH grant ran out. confirmedas of 2025-07-21

    23. [23]

      STAT reported in July 2025 that Google DeepMind's undisclosed one-time gift would support CASP for about 12 months, replacing NIH funding of around $639,000. reportedas of 2025-07-21

    24. [24]

      The CASP17 site lists the US National Institute of General Medical Sciences as sponsor, with support from Google DeepMind. confirmedas of 2026-10-10

      • CASP17 home page · Protein Structure Prediction Center (UC Davis) (retrieved 2026-10-10)
    25. [25]

      The Nobel committee described protein structure prediction as a 50-year-old problem that the AlphaFold AI model solved. confirmedas of 2024-10-09

    Revision history (2)
    1. Page created.
    2. Added the database's October 2026 size and curated collections, AlphaFold 3 access terms, open alternatives (Boltz, OpenFold3) and CASP funding details.

    Created Oct 10, 2026. Last reviewed by an editor on Oct 10, 2026. Next scheduled review: Jan 10, 2027.

    Cite this page

    "AlphaFold." ContentLora, updated Oct 10, 2026. https://contentlora.com/wiki/alphafold

    Spotted an error? Suggest a correction or emailcorrections@contentlora.com.

    Keep exploring

    Connections
    Google DeepMind Developed byNVIDIA Database partner in 2026 …Protein Data Bank Trained on data fromRFdiffusion Related design modelIsomorphic Labs Led by an AlphaFold Nobel…
    ☰ Contents