How Orikata Bio launched Noto on X
A world model for biology, with tools to learn from evolution and engineer biology
What Orikata Bio launched
How LaunchVideo.com rated the launch
LaunchVideo.com rating
About the launch
Orikata Bio is for scientists who search, compare, and design proteins, and for the models and agents now sitting in that workflow. The San Francisco company is a public benefit corporation in Y Combinator's Fall 2026 batch. Under the Orikata Bio brand it is training Ibuki, models that learn how sequence, structure, and function relate in evolutionary data, with protein folding as the first target. Megure Labs, the second brand, builds the training and inference software, including Noto, a workspace where people and agents can run analyses and hosted scientific models without provisioning GPUs or maintaining each model's dependencies. This launch is the first public release of that stack. Hikoboshi, for sequence and structure alignment, search, clustering, and evolutionary analysis, and Orihime, a differentiable alignment library used to train those comparisons, are open source. The company is also opening a web toolkit, a hosted inference preview in Noto, and access to Umi and Sora, collections it says contain more than 100 billion unique protein sequences and nearly 2 billion protein structures. Orikata's case for the timing is practical. Search, clustering, and alignment still feed structure prediction and design models, and the tools most labs use were written for much smaller databases. Limited free usage is available, with individual and team plans and paid model runs. The full multiplayer notebook is still in development. Casey Mogilevsky is founder and CEO. He is a former University of Pennsylvania MD/PhD student. At UC Berkeley he was jointly advised by Matt Francis and Jennifer Doudna, and that work produced publications and a patent underlying Catena Bio. At Cambridge he initiated a collaboration with David Baker's lab that later became part of Archon Bio. He is first author of an ICML 2026 paper on structured attention, the research Orihime is built on.
Tags
Questions
Frequently asked questions
The Orikata Bio launch post reached 4.5K views on X, with 26 likes, 6 reposts, 4 comments, 8 bookmarks.
Orikata Bio launched on X on November 16, 2026 with the message "A world model for biology, with tools to learn from evolution and engineer biology".
Orikata Bio is for scientists who search, compare, and design proteins, and for the models and agents now sitting in that workflow.








