All Competitors

Reference-guided anatomical segmentation for medical images, pairing vision-language spatial reasoning with a co-trained SAM 2 mask decoder.

Spectroscopy-grounded molecular foundation model that reads NMR, IR, and mass spectra as text, elucidating structures and generating 3D conformers.

Single-cell foundation model domain-adapting Llama-3.1-8B on 1.3M gastric cancer cells with gene-family cell sentences instead of ranked-gene order.

Missense variant pathogenicity predictor that also ranks candidate diseases, aligning ESM-2 protein embeddings with PubMedBERT disease text.

EHR foundation model for structured OMOP timelines that unifies patient embeddings, zero-shot outcome prediction and synthetic record generation.

Biological reasoning language model that post-trains Qwen3 with reinforcement learning from verifiable rewards for drug-discovery tasks.

Reasoning language model post-trained on virtual cell simulations, answering questions about gene perturbations and their effects in natural language.

Protein function prediction via compressed in-context learning on a sequence-structure language model, cutting 751-token demonstrations to under 16.

Medical image grounding model that localizes text phrases in CT, MRI, X-ray, ultrasound, endoscopy, dermoscopy, and fundus images.

Mammography report generation model, a LoRA adaptation of MedGemma-4B-it that writes narrative radiology reports carrying BI-RADS assessments.

Retinal OCT vision-language model that writes layer-by-layer clinical summaries and assigns six-class disease labels from a single B-scan.

Molecular reasoning language model for molecule captioning and text-to-molecule generation, trained by chain-of-thought distillation then reward RL.

Molecular structure elucidation model that reads IR, Raman, UV-Vis, NMR, and mass spectra as text and generates SMILES end to end.

Biomedical named entity recognition transformers, with task-specialized checkpoints for chemicals, diseases, genes, proteins, species, and anatomy.

Scientific multimodal foundation model, a 241B-parameter MoE with a tokenizer that reads molecular formulas and protein sequences natively.

Drug-drug interaction event text generation from two molecular structures, conditioned on biological functions selected for each drug in the pair.

Open medical multimodal models from Google, built on Gemma 3 with a medically tuned SigLIP vision encoder for clinical text and image understanding.

Multimodal foundation model for cardiac biosignals, pretrained by masked modeling on ECG, PPG, and clinical text from ~1.7 million individuals.

Biomedical vision-language assistant for medical visual question answering, pairing Phi-2 with a vision encoder in a 4.2B-parameter model.

Plasmid characterization and retrieval model aligning DNA sequences with property text across ten facets, from antimicrobial resistance to host range.

Multi-scale ECG-language model that aligns 12-lead ECG signals with clinical text at token, beat, and rhythm levels for zero-shot cardiac diagnosis.

CLIP-based vision-language foundation model for eye imaging, enabling zero-shot disease detection and cross-modal retrieval across 11 modalities.

Text-guided protein editing framework with disentangled structure and function latents, edited by rewriting either description at inference time.