ByteDance has raised $290 million for its artificial intelligence drug-discovery business after spinning the unit out of the Chinese internet giant, giving the fledgling company a $1.5 billion valuation and providing fresh capital for its expansion.

Shanghai-based Anew Labs completed its first external fundraising round, two people familiar with the matter told Reuters. The financing was led by HSG, formerly known as Sequoia China, IDG Capital and Hillhouse Investment, alongside 5Y Capital as a co-lead investor.

Other investors included Gaorong Ventures, Primavera Venture Partners and Boyu Capital, as well as strategic investor SBP Group and the state-backed Shanghai Future Industries Fund, the sources said. They spoke on condition of anonymity because the information has not been made public.

The financing marks a significant step for a company that was only recently separated from one of China’s largest technology groups. Anew Labs uses AI in drug discovery, a field where technology companies and investors are increasingly betting that machine learning can accelerate the identification of drug candidates and reduce some of the time and cost associated with conventional pharmaceutical research.

Anew Labs’ $1.5 billion valuation gives the business a substantial standalone market value despite its short history as an externally financed company. The fundraising also provides a new ownership structure under which ByteDance will retain a 56% stake, according to one of the sources.

The decision to spin out the unit reflects the different economics and operating requirements of AI drug discovery compared with ByteDance’s core internet businesses.

One source said ByteDance separated the AI drugmaker to better support its long-term development because AI-driven drug discovery follows a different industry logic and requires a different management approach from the group’s core operations.

Consumer internet businesses can often scale software products rapidly once they gain users, while drug discovery operates on a much longer development cycle and requires significant investment before commercial returns can emerge.

AI drug discovery also sits at the intersection of two capital-intensive industries. The technology requires substantial computing resources and specialized scientific talent, while successful drug development ultimately requires laboratory research, validation and, depending on the program, lengthy clinical development.

The new funding therefore gives Anew Labs the financial runway to develop its technology independently while retaining ByteDance as a majority shareholder.

ByteDance Builds an AI Portfolio Beyond Consumer Internet

The deal illustrates how Chinese technology companies are increasingly exploring AI applications outside the consumer products that made them global businesses.

ByteDance remains best known for TikTok and its Chinese counterpart Douyin, but the company has invested heavily in artificial intelligence across areas including foundation models, applications and infrastructure. Spinning out Anew Labs allows it to maintain exposure to another potentially valuable AI market while giving the drug-discovery business access to outside investors with expertise and capital specific to the sector.

The investor lineup is notable because it combines major Chinese venture and private-equity firms with a state-backed fund and a strategic investor.

That mix could provide Anew Labs with access to both financial resources and industry networks as it develops its technology. It also highlights the broader interest in AI as China seeks applications that extend beyond consumer-facing chatbots and into scientific research and industrial sectors.

The $1.5 billion valuation, however, should not be interpreted as evidence that AI has already transformed the economics of drug development. Drug discovery remains a high-risk business in which technical advances do not necessarily translate into successful medicines.

The value of an AI drug-discovery company ultimately depends on whether its models can identify promising compounds more efficiently, improve the probability of successful development, and translate those discoveries into commercially viable treatments. That makes the distinction between AI capability and pharmaceutical outcomes particularly important.

Anew Labs is entering a market where the potential payoff from successful drug candidates can be substantial, but where timelines are measured in years rather than the rapid product cycles common in consumer technology. The company’s ability to turn its AI capabilities into validated drug candidates will therefore matter more than the size of its initial funding round.

For ByteDance, retaining 56% means the company continues to have significant exposure to that upside while sharing the financing burden and bringing external investors into the business.

The structure could also give Anew Labs greater independence in attracting talent and capital. AI drug discovery requires expertise spanning machine learning, computational biology, chemistry, and pharmaceutical development, making it different from the engineering and product-management structures associated with ByteDance’s core platforms.

The fundraising is therefore as much about corporate structure as it is about capital.

ByteDance has effectively created a standalone vehicle through which investors can finance an AI application with a distinct development cycle, while the parent company preserves majority control.

For the wider Chinese AI industry, the deal adds to evidence that investment is moving beyond general-purpose models toward specialized applications where AI can potentially create value in sectors with high research costs.

Meta is expanding its subscription business across Facebook, Instagram and WhatsApp as the company looks to turn its growing investment in artificial intelligence into a new source of recurring revenue.

The company on Tuesday introduced Meta One, a collection of subscription plans offering expanded access to AI tools and premium features across its major social platforms. The plans give users greater access to tools for generating and editing images and videos, along with AI-powered features such as Instagram’s Restyle editing tool.

The move represents a broader monetization push for Meta’s AI ambitions. Rather than relying solely on advertising to recover the cost of developing expensive AI models and infrastructure, the company is now asking users, creators, and businesses to pay directly for greater access to those capabilities.

Meta’s AI strategy has expanded substantially since its $14.3 billion investment in Scale AI in 2025, which brought Scale AI CEO Alexandr Wang into Meta to lead its AI efforts. The new subscriptions provide a potential commercial channel for those investments by putting Meta’s Muse AI models directly behind paid tiers.

The announcement follows Meta’s introduction of separate subscription offerings for its social platforms in March. Instagram Plus and Facebook Plus cost $3.99 a month, while WhatsApp Plus costs $2.99, giving subscribers features such as profile customization, super reactions and story insights.

Early revenue data suggests those lower-priced offerings are gaining traction.

Market intelligence firm Appfigures estimates that Instagram’s average daily worldwide revenue reached $1.2 million during the week of Sept. 9, up 475% from the previous week. Facebook’s daily revenue reached $528,000, representing a 143% increase over the same period.

The figures suggest Meta may have found a potentially significant new monetization channel outside its advertising business, although the sharp week-to-week increases also reflect the relatively recent introduction of the subscriptions.

Meta targets AI power users, creators and businesses

Meta One’s consumer offerings are centered on two plans: Core at $7.99 a month and Premium at $19.99.

Both include the existing Facebook Plus, Instagram Plus, and WhatsApp Plus benefits, while their main differentiator is expanded access to Meta’s AI tools. Premium subscribers receive more AI usage than Core subscribers, allowing them to make greater use of features including Muse Image and Muse Video.

Subscribers can also use AI to edit Instagram Stories through Restyle and gain increased access to voice effects and other creative tools. The structure effectively creates a new premium layer around Meta’s social platforms. Users who want to use AI occasionally can continue with free tools, while those generating images, videos, or other content more frequently have an incentive to move to paid plans.

Meta is also targeting businesses and creators with substantially more expensive subscriptions.

Its business and creator lineup begins with Essential at $14.99 a month and rises through Advanced, starting at $49.99, Expert at $149, and Max at $499. Essential provides tools for managing a creator or business presence, expanded access to Meta Business Agent for responding to customers, a verified badge and channel on WhatsApp Business, as well as impersonation detection.

Advanced adds scheduling for Stories up to 30 days in advance, links in organic posts and Reels, exportable analytics, deeper audience insights, team-member access, additional linked devices, more business broadcast credits, and greater access to Meta Business Agent responses.

Meta is also introducing enhanced profiles for Essential and Advanced users that can highlight websites, locations and reviews. Reels will receive a more prominent follow button, while people who interact with content can automatically receive follow invitations.

The company plans to add more agent capabilities over time, including tools designed around marketing, business operations, content creation, and optimization. It will also introduce Edits Plus for Meta’s Edits content-creation app. The subscription will provide additional cloud storage for syncing projects across devices and increased access to the Edits AI assistant.

Expert and Max customers will receive the highest levels of feature access and Meta Business Agent capacity.

While the complex collection of plans could make Meta’s subscription strategy difficult for consumers to navigate, the underlying commercial objective is to create multiple payment tiers based on how heavily users depend on Meta’s platforms and AI services.

Subscription Push Could Become A Meaningful Revenue Stream

The early performance of Meta’s cheaper social subscriptions is giving investors a reason to take the strategy seriously.

Appfigures said U.S. users accounted for 32% of Instagram’s revenue and 39% of Facebook’s revenue after the subscriptions gained traction on Sept. 9, roughly 10% above each platform’s historical U.S. baseline.

The potential revenue opportunity becomes much larger if Meta can persuade users to pay not merely for cosmetic social features but for access to increasingly valuable AI capabilities.

BNP Paribas forecasts that Meta’s subscription push could add $13.5 billion in revenue by 2028, while Truist estimates the company could generate an additional $20 billion by 2030.

Those forecasts would still represent only a portion of Meta’s overall business, but they point to an important shift in the company’s monetization model. Advertising remains the economic engine of Facebook and Instagram, yet subscriptions could give Meta a second recurring revenue stream while simultaneously helping it monetize the costly AI systems it is building.

The approach is needed because AI infrastructure requires significant ongoing spending rather than a one-time investment. Meta therefore needs its AI capabilities to generate economic value, either by improving advertising, increasing engagement, or creating direct revenue.

Meta One addresses the third option.

The bigger test will be whether consumers see enough value in AI-generated content and productivity features to pay monthly fees when many competing AI tools are available separately. For businesses and creators, the calculation may be easier if Meta’s agents can demonstrably save time, improve customer response rates, or generate more revenue.

However, Meta’s subscription expansion is seen as an attempt to turn the company’s AI infrastructure from a largely cost-intensive investment into a direct consumer and business product, while deepening the commercial relationship between Meta and the users who already spend much of their time inside its ecosystem.