Alphafold

What Alphafold does

AlphaFold is an AI system developed by Google DeepMind that predicts protein three-dimensional structures from amino acid sequences with accuracy competitive with experimental methods such as X-ray crystallography and cryo-EM. The system uses deep learning neural networks in AlphaFold2 and diffusion model architectures in AlphaFold3, with the latter extending capabilities to protein-ligand docking, nucleic acid interactions, and complex molecular systems. AlphaFold2 was top-ranked in the CASP14 competition by a large margin and is credited as Nobel Prize-winning technology.

The system is delivered through two primary vehicles: the AlphaFold Protein Structure Database, hosted and operated in partnership with EMBL-EBI, which provides open access to over 200 million predicted protein structures with search, download, and API functionality; and open source code on GitHub for users needing custom predictions including multimer support. The portfolio also includes AlphaMissense, a sub-product for proteome-wide missense variant effect prediction. The most recent May 2026 release added approximately 31 million protein complex predictions across 4,777 proteomes, developed in collaboration with EMBL-EBI, NVIDIA, and Seoul National University.

AlphaFold operates as a free public resource under a CC-BY-4.0 license, with no direct revenue generated from the database itself. The product-led growth motion targets academic researchers, pharmaceutical and biotech companies, and drug discovery researchers globally, with distribution occurring through the EMBL-EBI web platform, a public Python API, open source distribution, academic publications, and training courses. As a non-commercial product within Google DeepMind, AlphaFold generates strategic value for Alphabet through research reputation, talent attraction, and indirect drug discovery partnership opportunities rather than direct revenue capture.

Name

Alphafold

Legal name

AlphaFold

Website

https://alphafold.ebi.ac.uk

Company type

Private

Founded year

2020

Operating status

Operating

Short description

AlphaFold is an AI protein structure prediction system developed by Google DeepMind, distributed free via the EMBL-EBI-hosted database. It serves academic researchers, pharmaceutical and biotech companies, and drug discovery researchers globally with 200M+ predicted structures.

Ownership category

akta.pro rank

Product category

AI Protein Structure Prediction Software

NAICS

Software Publishers (513210), Computer Systems Design and Related Services (54151)

SIC

Services-Computer Programming, Data Processing, Etc. (7370), In Vitro & In Vivo Diagnostic Substances (2835)

akta.pro primary industry

Protein Engineering & Directed Evolution Platforms (enzyme/therapeutic protein optimization) (HLAAAIAC)

akta.pro secondary industry

Engineered Protein Therapeutics (de novo/AI-designed proteins, novel scaffolds) (HLAAAAAJ)

Keywords

• Protein structure prediction
• AI protein modeling
• Computational biology
• Structural bioinformatics
• Drug discovery AI

Headquarters

HQ city

Hinxton

HQ country

United Kingdom

HQ region

Europe

Offices2 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Infrastructure

Revenue model

1. Free Public Access Database: AlphaFold Protein Structure Database is provided free of charge for academic and commercial use under CC-BY-4.0 license. No direct revenue is generated from the database itself.

Pricing tiers

Go-to-market motion1 record

Distribution channels2 records

Marketing channels4 records

Core offering

AlphaFold is an AI system developed by Google DeepMind that predicts a protein's 3D structure from its amino acid sequence with accuracy competitive with experimental methods. It is delivered as the AlphaFold Protein Structure Database — a free public database hosted by EMBL-EBI providing open access to over 200 million protein structure predictions — alongside open-source code on GitHub for custom predictions including multimer capability. The portfolio also includes AlphaMissense for proteome-wide missense variant effect prediction.

Product overview

AlphaFold is a protein structure prediction AI system developed by Google DeepMind, offered through two primary vehicles: (1) the AlphaFold Protein Structure Database, hosted by EMBL-EBI, providing open access to over 200 million predicted protein structures with search, download, and API functionality; and (2) the AlphaFold open source code available on GitHub for users who need to generate custom predictions beyond database coverage. The portfolio also includes AlphaMissense for missense variant effect prediction. AlphaFold regularly achieves accuracy competitive with experimental methods and was top-ranked in CASP14 by a large margin.

Differentiator

Problem solved

Functional benefit

Products and services

• AlphaFold Protein Structure Database A free public database hosted by EMBL-EBI providing open access to over 200 million protein structure predictions covering UniProt proteomes. Offers search, browse, download, and API access for academic and commercial researchers globally. Includes individual downloads for the human proteome and proteomes of 47 other key organisms important in research and global health, plus a Swiss-Prot curated subset.
• AlphaFold Open Source Code Open source implementation of AlphaFold available on GitHub that allows researchers to generate their own custom protein structure predictions, including support for multimer prediction of protein complexes. Free for academic and commercial use.
• AlphaFold AI Prediction System The core AI system developed by Google DeepMind that predicts a protein's 3D structure from its amino acid sequence with accuracy competitive with experimental methods. The system uses deep learning neural networks and includes AlphaFold2 and AlphaFold3 models.
• AlphaMissense AI model for accurate proteome-wide missense variant effect prediction, helping researchers identify potentially pathogenic genetic variants. Copyright 2023 DeepMind Technologies Limited.

Quantifiable outcome

• AI-assisted processes can reduce early-stage discovery timelines to under 2 years compared to traditional 10-15 year development cycles
• +2 more outcomes

Named customers1 record

Segments3 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

AI capability4 records

Feature5 records

Partnerships

Three partnerships are on record, tiered core and flagship.

• NVIDIAcoreTechnology or Integration · 1 March 2026March 2026 collaboration between EMBL-EBI, Google DeepMind, NVIDIA, and Seoul National University resumed work on protein complex prediction, undertaking a large-scale study of ~19M homomeric and ~8M heteromeric protein complex predictions across 4,777 proteomes.
• Seoul National UniversitycoreStrategic or Co-development Partner · 1 March 2026March 2026 collaboration on protein complex prediction undertaking large-scale study across 4,777 proteomes, including 16 model organisms and 30 WHO Global Health Proteomes initiative proteomes.
• EMBL-EBI (European Bioinformatics Institute)flagshipStrategic or Co-development Partner · 1 January 2021Google DeepMind and EMBL-EBI partnered to create the AlphaFold Protein Structure Database, making protein structure predictions freely available to the scientific community. EMBL-EBI hosts and operates the database, providing global access through their infrastructure. This is the primary partnership enabling worldwide distribution of AlphaFold predictions.

Scale indicators5 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• RoseTTAFold (Baker Lab / Institute for Protein Design): RoseTTAFold is the closest academic competitor to AlphaFold, providing deep-learning-based protein structure prediction from sequence. It pioneered three-track neural network architectures and is widely used by the same research community that relies on AlphaFold. RoseTTAFold-AllAtom extended capabilities to small molecules and nucleic acids, directly competing with AlphaFold3.
• ESMFold (Meta AI): ESMFold is Meta AI's protein structure prediction model that uses large-scale protein language models (ESM-2) rather than multiple sequence alignments, enabling faster inference. It directly competes with AlphaFold2/3 in providing open-access structure prediction at scale and has been benchmarked as the primary open-source alternative.
• Isomorphic Labs: Isomorphic Labs is Alphabet's commercial drug discovery subsidiary, spun out to commercialize AlphaFold technology for pharmaceutical partnerships. It is the closest commercial peer and a sister entity, applying AlphaFold-derived models to partner with pharma companies on drug discovery programs.

Emerging players

• Boltz-1 (MIT Jameel Clinic): Boltz-1 is an open-source protein structure prediction model developed at MIT, designed as a fully open and reproducible competitor to AlphaFold3. It targets biomolecular complex prediction including proteins, nucleic acids, and ligands, representing an emerging academic alternative with capability overlap to AlphaFold3.
• Chai Discovery: Chai Discovery is an emerging AI company focused on protein structure prediction and molecular design for drug discovery. It competes with AlphaFold in the AI protein structure and design space, particularly targeting antibody and therapeutic protein engineering applications.
• Cradle Bio: Cradle is an AI-driven protein engineering platform focused on designing improved proteins for therapeutics and industrial applications. It operates in the same AI protein design space as AlphaFold's extensions, with emphasis on generative protein engineering rather than structure prediction per se.
• EvolutionaryScale: EvolutionaryScale is an emerging AI company developing frontier protein language models (ESM-3) for generative protein design. It emerged from Meta's ESM research team and represents a next-generation competitor to AlphaFold with explicit de novo protein design capabilities.

Broad incumbents

• Recursion Pharmaceuticals: Recursion is a publicly traded AI-driven drug discovery company using high-throughput biology and machine learning. While not a direct protein structure prediction competitor, it operates in the adjacent AI drug discovery space where AlphaFold's structural biology outputs are increasingly applied to target identification and hit generation.
• Insitro: Insitro is an AI-driven drug discovery company combining machine learning with high-throughput biology for pharmaceutical R&D. It represents an adjacent player in AI-enabled drug discovery where AlphaFold-derived protein structure predictions are increasingly integrated into target validation and lead identification workflows.

Regional players

• Owkin: Owkin is a French AI biotech company applying machine learning to biomedical research and drug discovery. It serves as a regional and thematically adjacent peer, operating in AI-driven biology where protein structure models like AlphaFold are increasingly important for target discovery and translational research.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks5 records

Key highlights7 records

Customer concentration

• Xtwitter.com/GoogleDeepMind
• GitHubgithub.com/deepmind/alphafold
• LinkedInlinkedin.com/company/ebi

Compliance1 record

Revenue estimate

Valuation estimate

Number of profiles

Funding overview

Funding rounds

Investors

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Alphafold