All Competitors

Spatial omics foundation model that represents tissue as a hierarchical graph of neighboring cells over per-cell gene co-expression networks.

Ligand-binding protein design driven by a natural-language function description plus a ligand SMILES string, in 1B and 3B parameter variants.

Inverse folding framework combining a Markov bridge generator with direct preference optimization to design low-energy sequences and predict ΔΔG.

Spatiotemporal foundation model that learns representations directly from 4D functional MRI volumes for disease diagnosis and phenotype prediction.

Protein fitness prediction with end-to-end differentiable homology search, replacing MSA construction with vector search over 62M UniRef50 sequences.

EEG foundation model pairing a decoupled time-frequency tokenizer with a multi-scale state-space encoder for generalization under distribution shift.

Generative chemistry foundation model pairing a text-and-SMILES language backbone with a 3D molecular point-cloud encoder for prediction and design.

Sensor-language foundation models aligning wearable biosignals with text for zero-shot activity recognition, retrieval, and sensor captioning.

Protein language model trained with masked diffusion, unifying representation learning and generative design in one 650M-parameter model.

Inverse folding model refined by online reinforcement learning against folding and stability rewards, cutting design failure rates by 36-48%.

Virtual drug screening from per-atom protein and ligand embeddings retrieved by nearest neighbors. 30.4 EF1% on DUD-E at ~14 s per million molecules.

Protein function annotation model predicting Gene Ontology terms with direct preference optimization layered on a frozen ESM-C sequence encoder.

Diffusion model that backmaps coarse-grained protein structures to all-atom detail, scaling to condensates of over a million residues.

Generalist medical multimodal LLM for image understanding, visual question answering, and report generation across twelve-plus imaging modalities.

Multimodal ECG model that reads raw 12-lead waveforms and ECG images in one shared space via a structure-aware discrete signal tokenizer.

Structure-based virtual screening model that scores ligands against apo and predicted pockets, lifting blind-apo EF1% on DUD-E from 11.75 to 37.19.

Multimodal large language model that writes free-form natural-language gene function descriptions directly from a nucleotide sequence and a prompt.

Generative histopathology foundation model: a diffusion transformer trained on 30M H&E tiles, conditioned on self-supervised slide embeddings.

Domain-level language model treating Pfam protein domains as tokens to predict and design bacterial and fungal biosynthetic gene clusters.

RNA-protein contact prediction from sequence, built on ERNIE-RNA and ESM-2 embeddings. Reaches 0.77 auROC where AlphaFold 3 reaches 0.61.

Relative protein-ligand binding affinity prediction from docked complexes, matching Schrodinger FEP+ ranking accuracy zero-shot on the FEP benchmark.

Spatial proteomics foundation model, marker-aware and panel-agnostic, pretrained on 47 million multiplexed tissue-imaging patches from 175 markers.

Codon optimization model for heterologous expression in E. coli, fine-tuning ProtBert to label each residue with an expression-weighted codon.