Sophont builds open, universal medical AI that understands pathology, neuroimaging, clinical text and more—empowering clinicians and researchers worldwide.
Medmarks, our LLM benchmark for medical tasks →
Collaborate with us at the frontier of open medical AI. Whether you're a healthcare provider, research institution, or investor, let's connect.
Sophont (blog), 2025
Launches the Medmarks evaluation suite with 20 verifiable and open-ended medical benchmarks, 46 models across 56 configurations, and RL-ready environments to track LLM medical capabilities.
Sophont (blog), 2025
We open-sourced OpenMidnight, a replication and improvement of Kaiko’s Midnight model that reaches state-of-the-art eva + HEST benchmark performance with only 12K TCGA slides, $1.6K of compute, and a fully public training pipeline.
NeurIPS workshop, 2025
We project volumetric fMRI into videos of 2D flat-map activity and pretrain spatiotemporal MAE Vision Transformers on 2.3K hours of Human Connectome Project data, revealing strict power-law scaling and strong downstream decoding of both shared brain states and subject-specific traits.
NeurIPS (spotlight), 2023
Introduced a novel method to reconstruct seen images from fMRI brain signals by aligning brain activity with image embeddings and a diffusion model.
ICML, 2024
Demonstrates that a shared fMRI-to-image model can be fine-tuned on just one hour of data from a new person, enabling high-quality brain decoding with minimal data.
Nature Biomedical Engineering, 2024
Presents RoentGen, a model that can generate realistic, high-resolution X-ray images from clinical text prompts, aiding in data augmentation and education.
AAAI, 2024
This model generates preliminary reports from chest X-rays, significantly improving radiologists' diagnostic efficiency without compromising accuracy.
Blog post, 2025
Provides a comprehensive overview of how LLMs are evaluated for medical applications, highlighting the critical need for more robust, multimodal evaluation frameworks.
Blog post, 2021
Details the community-led creation of DALL·E mini, an open-source text-to-image model, explaining its architecture, training, and impact.
JOSE, 2021
A lightweight, open-source JavaScript library for visualizing 3D brain maps directly in a web browser to share and explore fMRI meta-analysis results.
Optica, 2023
A microscopy technique using deep learning to create virtual H&E stains of unlabeled tissue, providing real-time histology without chemical staining.
arXiv, 2024
This study explores using fNIRS, a portable alternative to fMRI, for decoding visual imagery, showing promise in classifying imagined visual categories.
ICML, 2024
Introduces the Hourglass Diffusion Transformer (HDiT), a novel architecture for generating high-resolution images directly in pixel space.
Nature Communications, 2024
Identifies key challenges in NeuroAI training and provides recommendations for fostering an interdisciplinary environment for the next generation of scientists.
CVPR, 2025
Introduces NSD-Imagery, the first large-scale fMRI dataset dedicated to mental imagery, designed to advance brain decoding models for imagined visuals.
Under review
Presents a robust multimodal architecture that adapts fMRI-to-image models from visual perception to mental imagery, improving decoding of imagined scenes.
Company Blog, 2025
We closed $9.22M across pre-seed and seed rounds led by Kindred Ventures to accelerate universal multimodal medical foundation models and expand the MedARC community.
Interactive Article, 2025
Our MindEye2 paper is discussed in a NY Times article exploring AI's remarkable capabilities across different body parts, specifically in the Brain section.
Blog Post, 2025
Tanishq shares his Medical AI Manifesto, introducing Sophont and our mission to build open multimodal foundation models for healthcare.
Media Feature, 2025
Media collaboration with Fujifilm Australia showcasing our breakthrough technology for reconstructing mental imagery from brain signals.
Podcast Episode, 2025
Cognitive Revolution podcast where Tanishq discusses the future of brain-computer interfaces and medical AI.
Podcast Episode, 2025
Cognitive Revolution podcast where Paul dives into how we can decode thoughts from shared latent spaces, and the technical aspects of brain decoding.
MIT Presentation, 2024
Tanishq presents at MIT about open medical foundation models and their transformative potential for the future of healthcare.
Congressional Hearing, 2023
US Senate hearing briefly discusses our work as an example of medical AI technology applications in healthcare.
Podcast Episode
Jeremy Howard and Tanishq go in-depth into Tanishq's upbringing, life journey, and path to becoming a leading AI researcher.
TV Appearance
Young Tanishq appears on Conan O'Brien's show as a child prodigy and shares some science jokes.
2025
2025
2025
2025
2024
2023
2023
We are fortunate to partner with investors and strategic allies across medicine, AI research, and venture—including founders of iconic developer tools, healthcare funds, and leaders from Google DeepMind, Hugging Face, and more.
Steve Jang
Kindred Ventures
Kanyi Maqubela
Kindred Ventures
Mei Zuo
Mei Ventures
Jeff Dean
Google Chief Scientist
Clem Delangue
Hugging Face CEO via Factorial Capital
Matthew Hartman
Factorial Capital
Lukas Biewald
Weights & Biases co-founder
Logan Kilpatrick
Google DeepMind
Delphi Ventures
Matt Miller
Headwater Ventures
fal
Waikit Lau
Brightwing Ventures
Kevin Zhang
Upfront Ventures
Mukaya (Tai) Panich
Aiconic Ventures