Benchmarks

Every model, measured on both sides

What TT-Bio runs, how fast each model is on Tenstorrent and on NVIDIA, what that costs per prediction, and how every number was taken.

What TT-Bio runs

Structure and binding affinity prediction

--modelFoldsAffinityMSA
openfold3protein, RNA, DNAnoon by default
boltz2protein, DNA, RNA, ligandyeson by default
openddemulti-chain protein complexesnoon by default
protenix-v2protein, DNA, RNA, ligandnoon by default
rf3protein, nucleic acid, ligandnoon by default
esmfold2single protein chainnonot needed
openbindprotein, RNA, DNA, ligand by SMILES or CCDnoon by default
protenix-v1protein, DNA, RNA, ligandnoon by default
mol1protein, DNA, RNA, ligandnoon by default
af2iga designed complex, from its own coordinatesnonot needed
nesso1nothing, it returns scalarsyesnot needed

OpenFold3, Protenix-v2, Protenix-v1 and Mol-1 are AlphaFold3-family. AF2-IG scores a designed complex rather than folding a sequence. Nesso-1 runs under tt-bio affinity and returns scalars, not coordinates. esmfold2-fast is lighter, opendde-abag is the antibody-antigen checkpoint.

Protein design

--modelModelDesignsInput
rfd3RFdiffusion3all-atom binders, motif scaffolds, nucleic-acid bindersJSON spec with contig strings
boltzgenBoltzGenprotein, peptide, nanobody, antibody binderstarget structure in a design YAML
rfd4RFD4-Proteinabinders, motif scaffolds, enzymes, antibody and nanobody CDRs, unconditional monomersRFD4 design spec (YAML or JSON)
pxdesignPXDesignbinder backbones, no sequencetarget structure in a design YAML

Protein embeddings

--modelEmbeddingsOutputSizes, as --model
esmcsequence, ESMC language modelper-residue, pooledesmc-300m, esmc-600m, esmc-6b
saprotstructure-aware, ESM-2 over a fused amino-acid and Foldseek-3Di vocabularyper-residue, pooledsaprot-35m, saprot-650m, saprot-1.3b

SaProt's structure input is optional; sequence-only costs accuracy below 1.3B.

Performance and cost

Throughput per dollar of purchase price

Derived

Throughput per dollar of total cost of ownership

Derived

Predictions and designs per hour, per server

Derived

What the server costs

Price and power

One AI Processor the measured seconds behind every server figure

Prediction time, one AI Processor

Measured

Seconds per design or affinity prediction, one AI Processor

Measured

The exact figures every bar as a table

Measured, seconds per prediction on one AI Processor

    Measured, seconds per design or affinity prediction on one AI Processor

    Derived, predictions and designs per hour per server

    Derived, throughput per dollar of purchase price

    Derived, throughput per dollar of total cost of ownership

    Methods and sources what ran, how cost is computed, prices and power

    Sources

      Caveats

        Every seconds figure is measured on the hardware named. Everything else is derived from those measurements and the published prices and power ratings. Updated .

        Tenstorrent, Blackhole and Galaxy are trademarks of Tenstorrent Inc. NVIDIA, DGX, H200, B200 and A100 are trademarks of NVIDIA Corporation. Hardware names are used to identify the systems measured.