ByteDance closes first external funding round for AI drug discovery spin-off Anew Labs, raising 290 million dollars at a valuation of 1,5 billion dollars

ByteDance closes first external funding round for AI drug discovery spin-off Anew Labs, raising 290 million dollars at a valuation of 1,5 billion dollars

On September 16, 2026, TikTok owner ByteDance closed the first external funding round for its AI drug discovery business, Anew Labs, raising 290 million dollars at a valuation of 1,5 billion dollars while retaining a 56% stake. Investors included the Chinese state-backed Shanghai Future Industries Fund. Anew Labs is preparing an investigational new drug application for its most advanced candidate, a small molecule targeting the inflammatory protein IL-17. This application is a step toward human trials.

Anew Labs began within ByteDance in 2021 under Liu Kai, who had previously spent seven years as a venture investor at IDG Capital. In June 2026, it became an independent company, taking its algorithms with it. It has 50 researchers in Shanghai, San Jose and Singapore. A drug undergoes years of regulatory review, requiring a different approach to financing from an internet product that can be refined in weeks. ByteDance therefore established the business as a separate company with external shareholders, retaining a controlling 56% stake and supplying computing resources through its own cloud service, Volcano Engine.

The round was first reported by Reuters. It was led by HSG (formerly Sequoia China), Liu Kai’s former employer IDG Capital, and Hillhouse’s venture capital arm.

Anew Labs predicts protein structures using AnewFold and Protenix. Protenix is an open-source replica of Google DeepMind’s AlphaFold 3, whose creators received the 2024 Nobel Prize in Chemistry. The company’s most advanced candidate is a small molecule targeting the IL-17F homodimer, a pair of identical copies of the inflammatory signaling protein IL-17, which drives psoriasis and other chronic inflammatory conditions. Anew first presented this IL-17-targeting molecule at the Immunology2026 and ICLR conferences in May 2026. Approved IL-17 drugs are injectable antibodies. A small molecule could be developed into a pill, although the target had long been considered inaccessible to this class of compounds. The project remains at the preclinical stage.

The round places Anew Labs alongside Isomorphic Labs, Google DeepMind’s drug discovery spin-off, and Insilico Medicine, which already has candidates in human trials. All three advance their own molecules into clinical trials rather than selling a platform, giving them both a share of any success and exposure to the risk of failure. According to the South China Morning Post, Chinese AI drug discovery startups raised hundreds of millions of dollars that same week. Chinese capital is supporting domestic competitors to Anthropic and Isomorphic Labs.

In its 2026 report, the IQVIA Institute, a pharmaceutical industry data provider, described a “credible signal”: emerging biopharmaceutical companies using AI had an average Phase 1 success rate of 75%, substantially higher than comparable programs without AI. By Phase 2, the difference had almost disappeared. An August review in Nature Reviews Drug Discovery described the same gap from another perspective: methods for comparing AI tools have become plentiful, while evidence of actual clinical benefit remains much more limited. According to calculations by the consulting firm Deloitte, the average cost of developing a drug, including the cost of failures, rose to 2,671 billion dollars in 2025 from 2,229 billion dollars a year earlier.

The market values Anew Labs at 1,5 billion dollars. IQVIA’s data, however, support an AI advantage only at the least expensive stage of drug development. The IL-17F application will be one of the first tests of whether that advantage can hold through longer, more expensive phases.