Tahoe
Founded 2022 Β· 39 employees on LinkedIn Β· 5 known investors
Tahoe Bio AI applies machine learning and computational biology to drug discovery and development. The company leverages AI and advanced cellular modeling for pre-clinical research in the biotech industry.
Also known as Tahoe Bio Β· Tahoe Bio AI Β· Tahoe Therapeutics
Investors Β· 5
Company profile
researched Aug 2026Tahoe (also referred to as Tahoe Therapeutics) is a biotechnology company building large-scale biological datasets and machine learning systems intended to provide experimental verification for hypotheses generated by reasoning models. The company generates perturbative single-cell data across primary cells, organoids, iPSC-derived cells, and in vivo models with immune context, positioning this data as "ground truth" for biological AI.
On top of this data, Tahoe develops a family of machine learning models, including Rhaister, which learn the structure of the generated data to produce synthetic data for new biological contexts and predict cellular responses to perturbations not directly observed; the company states these models train in seconds and run in milliseconds. It also builds Tara, an autonomous research agent that traverses the dataset, generates hypotheses, validates them against the data, and surfaces candidate discoveries. Prior releases include the open-sourced Tahoe-100M dataset and the Tahoe-x1 model, alongside pre-clinical chemistry and drug discovery capabilities reflected in its team composition.
Founding story
Tahoe was founded by Nima Alidoust (CEO), Johnny Yu (CSO), Hani Goodarzi, and Kevan Shokat. Alidoust holds a Princeton Ph.D. and previously worked at Rigetti, McKinsey, and 1QBit; Yu holds a UCSF Ph.D. with prior experience at Biogen, Broad, and Millennium; Goodarzi holds a Princeton Ph.D. and is an associate professor at UCSF; Shokat is a professor at UCSF, UC Berkeley, and HHMI. The company describes its founding team as having discovered drugs for difficult targets, trained frontier AI models on chemical and biological data, and invented methods in single-cell and cancer biology.
Business model
Tahoe generates proprietary perturbative single-cell datasets and builds machine learning models and research agents on top of them. It releases datasets and models openly (for example Tahoe-100M and Tahoe-x1) and pursues drug discovery through partnerships and joint ventures, including a planned joint venture with Alloy Therapeutics to develop antibody-drug conjugates, and a data-generation partnership with the Arc Institute and Biohub.
Traction
Tahoe has open-sourced the Tahoe-100M dataset (February 2025) and released the Tahoe-x1 model (October 2025). It announced a partnership with the Arc Institute and Biohub in January 2026 to generate a large perturbation dataset for virtual cell models, and a planned joint venture with Alloy Therapeutics for antibody-drug conjugates. Its team page lists roughly 25 staff spanning biology, computational biology, machine learning, pre-clinical chemistry, and operations.
βΈFull profile β market position, technology, go-to-market, history
Market position
Tahoe operates in AI-driven drug discovery and virtual cell modeling, working alongside research institutions such as the Arc Institute and Biohub and biotech partners including Alloy Therapeutics.
Tahoe states that it generates perturbative single-cell data at a scale and diversity that no other organization can match, spanning primary cells, organoids, iPSC-derived cells, and in vivo models with immune context, and that its models are simpler yet outperform more complex alternatives while training in seconds and running in milliseconds. It couples this data with an autonomous agent that closes the loop between hypothesis generation and experimental verification.
Technology
Tahoe's platform combines high-throughput generation of perturbative single-cell data across primary cells, organoids, iPSC-derived cells, and in vivo models with immune context; a family of machine learning models including Rhaister that generate synthetic data for unseen biological contexts and predict cellular perturbation responses; and Tara, an autonomous research agent that queries the data, forms hypotheses, and validates them. Released artifacts include the Tahoe-100M dataset and the Tahoe-x1 model, and the company distributes work via GitHub and Hugging Face.
Go-to-market
Researchers and organizations working on computational biology, virtual cell models, and pre-clinical drug discovery, including academic research institutes and biotech partners.
History
The company, which also operates under the name Tahoe Therapeutics, open-sourced the Tahoe-100M dataset in February 2025 and introduced the Tahoe-x1 model in October 2025. In November 2025 it announced the addition of drug discovery veterans Wayne Spevak and Ben Powell to its leadership team. In January 2026 it announced a partnership with the Arc Institute and Biohub to generate a large perturbation dataset for virtual cell models, and a planned joint venture with Alloy Therapeutics focused on antibody-drug conjugates for hard-to-treat cancers. In June 2026 it introduced Rhaister and Emerald Bay.
Compiled by commissioned research from 8 cited public sources β announcements, filings, and press listed under research sources below.
Timeline Β· 6
launches, deals, and filingsTahoe announced Rhaister and Emerald Bay in a post authored by CEO Nima Alidoust.
Tahoe Therapeutics and Alloy Therapeutics announced plans to form a joint venture to develop first-in-class antibody-drug conjugates for hard-to-treat cancers.
Tahoe announced a partnership with the Arc Institute and Biohub to generate what the company describes as the largest perturbation dataset for virtual cell models.
Tahoe announced the addition of drug discovery veterans Wayne Spevak and Ben Powell to its leadership team; both are listed on the company team page in pre-clinical chemistry and drug discovery roles.
Dated company events from announcements, filings, and press; legal rows summarize public dockets and regulator releases.
In the news
βΈResearch sources Β· 8
primary sources listed
- Tahoetahoebio.ai Β· web
8 public sources were cited for this profile; the first-party ones are listed here.
Frequently asked questions
- What does Tahoe do?
- Tahoe generates large-scale perturbative single-cell data and builds ML models and research agents to verify biological hypotheses.
- Who are Tahoe's investors?
- Tahoe's investors include AIX Ventures, Amplify Partners, Civilization Ventures Management, Llc, Wing, Databricks Ventures.