AI for Bio · Model

AlphaGenome

Sequence-to-function model reading 1 Mb of DNA for base-resolution regulatory tracks and variant effects

Genomics & RNABig-tech labUnited KingdomSuperseded by a successor中文

Overview

NameAlphaGenome · aliases: AlphaGenome
Full nameThe developer uses the model name only
Ambiguous names (not auto-linked)AlphaGenome 2
FamilyNot a family entry. The successor AlphaGenome 2 (2026-11-11) is recorded as the latest event here; no family entry
OrganizationGoogle DeepMind
Organization typeBig-tech lab
CountryUnited Kingdom
Model classGenome sequence-to-function model / variant effect prediction
DomainGenomics & RNA
First release† partly verified
Basis: Developer blog announcement with API preview (preprint 2025-06-27)
Note: The announcement date was not directly re-verified and comes from aggregators source
Other milestones
  • First public evidence: † · Blog announcement and API preview source
  • Preprint v1: · bioRxiv v1 (registered date) source
  • Journal online: · Journal online publication source
  • Code release: · Inference code released with the journal paper source
  • Weights release: · Weights released under non-commercial terms source
  • Official announcement: † · Developer blog source
Latest update · successor
Basis: Successor model AlphaGenome 2 launch and API v2.0.0
AlphaGenome 2 launched: a conditional model predicting one tissue or cell type at a time, reported by the developers as trained on over a hundred million single-cell profiles; available by API for non-commercial use with a commercial private preview, weights not mentioned
Note: The published record did not include this event source
The most recent sourced event, not a claim that nothing changed since.
StatusSuperseded by a successor · The successor AlphaGenome 2 has been released
OpennessCode: Open · documentation and example code CC-BY-4.0
Weights: Non-commercial · non-commercial terms, acceptance required
Weights and inference code public from 2026-01-28; API open for non-commercial use; the successor's weights are not mentioned
Benchmarks reported3 counted (self 1, modified protocol 2)
Wet-lab validationNone recorded
Unverified fields† first_release, release_milestones.official_announcement, training_data.a_public
As of2026-11-16

What it does

Reads 1 Mb of DNA and predicts expression, transcription initiation, chromatin accessibility, histone marks, transcription-factor binding, contact maps and splice-site usage at base resolution, scoring non-coding variants by the difference between reference and alternate predictions. Compared with Borzoi it lengthens the input and merges modalities into one model; unlike self-supervised DNA language models such as Evo 2 and NTv3 it is supervised on functional-genomics tracks.

Benchmarks reported

Every row is one pairing from the survey's relation file; a score belongs to the benchmark page's ranking, where it is shown with metric, eval set, source type and date. 3 of 3 pairings are counted.

BenchmarkRelationEval set / protocolSource typeNote and evidence
08 Basenji/Enformer test setDeveloper report, official protocol—Self-reportedSelf-reported on this data's track test set · evidence
11 ClinVar hold-out protocolDeveloper report, modified protocol—Self-reportedIts own variant-effect suite, including splicing classes from this database †partly verified · evidence
37 TraitGymDeveloper report, modified protocol—Self-reportedUses the variant set and matched controls but reports enrichment rather than AUPRC, so it is not board-comparable. †partly verified · evidence

Other evaluations (outside the 44)

  • QTL and causality tasks (Own task set) Self-reported: Direction and coefficient correlation for expression and accessibility QTLs, plus a supervised classifier trained on multimodal scores, all developer-defined
  • Public reporter-assay datasets (Public benchmark not among the 44) Self-reported: Splicing and regulatory evaluations on two public MPRA datasets, outside the 44
  • Third-party probing comparison (2026-04) (Comparison against baselines only) Third-party: A third-party study claims probes trained on another genome model's embeddings beat this model; those are probes rather than zero-shot runs and are not directly comparable
  • BEND (Public benchmark not among the 44) Self-reported: The original paper does not report this benchmark

Developers report: one model matches or beats the strongest external model per task on most track and variant-effect evaluations, with multimodal prediction in a single 1 Mb context.

Independent evidence: No independent evidence found. No third-party evaluation under a site benchmark protocol was found even after the weights were released; existing third-party work uses a probing setup

Training data

Group in Figure 7: Public data only.

ClassValueDetail
a. Public databasesYesPublic functional-genomics and variant resources; the paper states no new raw experimental data was generated source
b. Private or licensed dataNot disclosed
c. Own wet-lab data used for trainingNo source
d. Wet lab used only for validationNoThe reporter-assay and variant validations in the paper use public data only source

Wet-lab validation

None. The developers report no wet-lab validation.

Background

  • Organization: A London AI lab inside a US-listed technology group, founded in 2010, acquired in 2014 and merged with another group research unit in 2023-04. Headcount framings differ: a 2025-09 media report cites about 5,600 (a wider remit) while a data firm counts 3,678 for its UK entity (2025).
  • Funding: Funded from the group's R&D budget; a sister AI drug-design company in the same group announced a $600M first external round on 2025-03-31.
  • Products and partnerships: A sibling co-folding model AlphaFold 3; two earlier sequence-to-function models, one of which (Borzoi) is both predecessor and comparator; at API launch the developers said thousands of scientists were testing it; the group also has the single-cell model C2S-Scale with a US university.
  • Related models: Borzoi, AlphaFold 3, C2S-Scale, Evo 2, NTv3

Reading notes

  1. All win counts are self-reported and measured against the strongest external model per task, not a rank on a common third-party board.
  2. The 2025-06 preview and the journal version count differently; always state which version is cited.
  3. The ClinVar side covers splice-related classes only and does not show general pathogenicity classification is usable.
  4. Long-range regulation (enhancer-gene links beyond 100 kb) is an acknowledged weakness, and only TraitGym's variant sets were used.

Caveats

  1. The preview and journal versions count differently and must not be mixed.
  2. Weights carry non-commercial terms, and the model card and paper conflict on whether one tissue dataset is in the released version.
  3. All site benchmark entries are self-reported; no third-party figures are citable.
  4. The successor is API-only, with weights and evaluation detail undisclosed.

Sources

  1. Journal paper (journal)
  2. Journal full text (public archive) (journal)
  3. Developer blog (preview) (company blog)
  4. Developer blog (successor model) (company blog)
  5. Code and weights repository (repository)

Corrections

2 corrections
FieldWasNowBasisFound
latest_updateNo 2026 event2026-11-11 successor model AlphaGenome 2 released (API v2.0.0)source2026-11-16
statusActiveSuperseded by a successorsource2026-11-16

Part of the AI for Bio benchmarks survey · all 112 models · open in the explorer. Record as of 2026-11-16.