AI for Bio · Model

Boltz-2

Open model that adds a small-molecule binding-affinity head on top of co-folding

Structure & co-foldingMixedUnited StatesMaintained only中文

Overview

NameBoltz-2 · aliases: Boltz-2
Full nameOfficial materials use Boltz-2 only; no expanded name
Ambiguous names (not auto-linked)Boltz
FamilyNot a family entry. Covers Boltz-2 only; Boltz-1/1x and BoltzGen have their own records
OrganizationMIT, Recursion, Boltz PBC
Organization typeMixed · Jointly released by a university lab and a listed biotech company, now maintained by a public benefit corporation
CountryUnited States
Model classCo-folding plus affinity prediction
DomainStructure & co-folding (also: Small molecules)
First release · code release
Basis: GitHub release v2.0.0; preprint 2025-06-18 source
Other milestones
  • Preprint v1: · Preprint v1 source
  • Code release: · GitHub release v2.0.0 source
  • Weights release: † · Weights published with v2.0.0 under MIT source
  • Official announcement: † · Joint release note by the institute and the partner company source
Latest update† unverified · product update
Basis: Product-line event: two new models and a hosted API went live (secondary sources only)
The developers announced two new models and a hosted API running this model; the open repository's last release was 2025-09-08 and the weights themselves are unchanged (unverified)
Note: The date appears only in secondary coverage; the developers' own page was not retrieved (unverified) source
The most recent sourced event, not a claim that nothing changed since.
StatusMaintained only · No open-repository release since 2025-09; the 2026 product event is unverified
OpennessCode: Open · MIT
Weights: Open · MIT
Weights and inference code are MIT-licensed; as of 2026-11-16 the repository still marks the training code and evaluation scripts as coming soon
Benchmarks reported3 counted (board 2, retrospective 1, cited only 1, unconfirmed 2)
Wet-lab validation1 recorded campaign
Unverified fields† latest_update, training_data.b_private_licensed, release_milestones.weights_release, benchmarks_reported[1], benchmarks_reported[2], benchmarks_reported[3], benchmarks_reported[5], background.products_partnerships
As of2026-11-16

What it does

On the structure side it adds method conditioning and more control over the predecessor; on the affinity side a head on top of the structure module outputs binder or non-binder plus binding strength for hit finding and hit-to-lead ranking. That affinity side is what separates it from Boltz-1/1x and Chai-1, and Nesso-1, TerraBind and AQAffinity reuse its public-data recipe.

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 6 pairings are counted.

BenchmarkRelationEval set / protocolSource typeNote and evidence
31 FoldBenchOfficial leaderboard or resultsmaintainer run on post-2024-01 subsetBoardMaintainer ran Boltz-2 on post-2024-01 subset (training cutoff later; reference only) †partly verified · evidence
34 AbBiBenchOfficial leaderboard or resultsregistered interface-score entryBoardBenchmark-registered ipSAE entry; not structure accuracy †unverified · no source link
01 CASPRetrospective dataset useaffinity track, retrospective (2 targets / 140 pairs)Third-partyRetrospective CASP16 affinity data (2 targets, 140 pairs); not a blind participation · evidence
30 PLINDER / PINDER / Runs N' PosesCited as comparison only (not counted)Runs N' PosesThird-partyQuoted in the Pearl paper and flagged not directly comparable; its official evaluation docs omit this benchmark †unverified · no source link
12 DUD-E / LIT-PCBAUnconfirmed (not counted)decoy-set screening in a third-party paperUnconfirmed sourceThird-party paper uses DEKOIS 2.0 etc., not clearly DUD-E/LIT-PCBA; second-hand †disputed · evidence
27 Polaris / ASAP / OpenADMETUnconfirmed (not counted)—Unconfirmed sourceSite page says it matched the antiviral top 5 without fine-tuning; no numbers; participation not openly verified †disputed · no source link

Other evaluations (outside the 44)

  • Free-energy calculation comparison subset (Own task set) Self-reported: Correlation against a physics-based method on a four-target subset, self-reported; scores are not shown here
  • Hit-to-lead and screening sets, self-reported (Own task set) Self-reported: Self-reported results on a public hit-to-lead dataset and a high-throughput screening dataset, neither a site benchmark
  • Third-party prospective evaluation (third party prospective evaluation) Third-party: A prospective study on about 557 co-crystal complexes finds its affinity score best for early enrichment while its poses trail the contemporary closed model
  • Third-party protein-protein affinity study (third party evaluation) Third-party: A paper finds it still below a sequence baseline after fine-tuning

Developers report: near physics-based free-energy accuracy on a four-target subset at about three orders of magnitude lower cost, matching the top entries of a public challenge without fine-tuning.

Independent evidence: Third-party prospective work finds the affinity score useful for enrichment but insufficient alone for virtual screening; other work argues screening results are inflated by training overlap and that it trails a sequence baseline on protein-protein tasks.

Training data

Group in Figure 7: Public data only.

ClassValueDetail
a. Public databasesYesPDB (to 2023-06-01), MD ensembles, distillation plus public affinity libraries such as ChEMBL; totals differ by source source
b. Private or licensed dataNot disclosedNone in the public version; from 2026-01 a pharma partner's historical data is used for refinement, with no statement on the public version
c. Own wet-lab data used for trainingNo
d. Wet lab used only for validationNoNo new wet lab of its own; a partner's eight internal assays serve as a held-out blind test and one prospective item is a simulation source

Wet-lab validation

Every rate carries its numerator and denominator. Claims that could not be verified are shown as unverified, without numbers; different paper versions are separate entries.

Partner blinded assays (journal paper)
Targets 8 · Designs tested — · Hits 3 · Success 3/8 assays above the stated correlation threshold · Best affinity —

Denominator basis: Eight internal partner assays, using existing experimental data as a held-out blind test rather than wet lab run for this model
Partner-lab executed and developer-compiled, so not an independent replication; the correlation threshold is a metric and is not shown here

partner-lab · · verified · source

Background

  • Organization: At release it was co-authored by two research units of a US university and a listed biotech company that supplied data and compute; it is now maintained by a public benefit corporation spun out of that university, which says it does not develop its own drugs.
  • Funding: The maintainer launched publicly on 2026-01-08 with a $28M seed round co-led by three firms; valuation undisclosed.
  • Products and partnerships: A platform product launched the same day, alongside a pharma collaboration building custom structure, affinity and biologics-design models on that company's data (custom development, per second-hand analysis); the two new models and hosted API of 2026-06-16 appear only in secondary coverage (unverified).
  • Related models: Boltz-1 / Boltz-1x, BoltzGen, Nesso-1, TerraBind, AQAffinity, IsoDDE

Reading notes

  1. The free-energy comparison holds only on a four-target subset, and only high-sequence-identity targets were removed; overlap with other evaluation sets is not reported set by set.
  2. Only a minority of the eight partner blind assays clear the threshold, so performance on new targets is unstable and the subset result cannot be extrapolated.
  3. Because of its later cutoff the structure side is often excluded from PoseBusters-style comparisons, so no claim either way about structure quality follows.
  4. The 2026 product event comes from secondary coverage, and the weights themselves are unchanged since 2025-09.

Caveats

  1. Training code and evaluation scripts are still marked coming soon in the repository, so full reproduction of the paper is not yet possible.
  2. The most recent update event has secondary sources only and must display as unverified.
  3. Three pairings rest only on the site's own pages (unverified).
  4. One pairing uses a different decoy set and its membership of that family is still undecided.
  5. The wet-lab item is a held-out blind test on a partner's existing data, neither new experiments nor an independent replication.

Sources

  1. The model's preprint (preprint)
  2. Paper full text (public archive) (paper)
  3. Official repository (repository)
  4. Official repository releases (repository)
  5. Institute release note (official site)
  6. Third-party prospective evaluation preprint (preprint)
  7. Third-party re-run paper (third-party paper)
  8. Coverage of the 2026 product event (second-hand) (second-hand) second-hand

Corrections

3 corrections
FieldWasNowBasisFound
latest_updateThe 2026-06-16 product event was presented as settled factMarked unverified: secondary coverage only, with the developers' own page not retrievedsource2026-11-16
other_evaluationsA free-energy correlation value relayed by a second-hand pageCorrected to the paper's own figure, and this record shows no concrete scoressource2026-11-16
wetlab_validationThe partner assays were treated as this model's own wet-lab validationReclassified as a held-out blind test on existing data, partner-lab executed, with one prospective item a simulation rather than an experimentsource2026-11-16

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