90,000 Litres Idle for Two Years: China's Biologics CDMO Capacity Shakeout Meets the AAV/LNP Funding Wave

At Ledu Life Science Park in Shanghai's Baoshan district, six 15,000-litre bioreactors — 90,000 litres of biologics capacity in total — have sat idle for more than two years, and the company's biologics subsidiary entered court-administered bankruptcy liquidation in December 2025. Ledu had once pitched a 1.3-million-litre capacity target in 2021, within reach of Samsung Biologics' 2032 plan. Meanwhile the capital has not left: vector-focused CDMOs — Yinuo, PackGene, Jinfan — closed consecutive nine-figure-yuan rounds within a year. The old moat has become an entry ticket, and next-generation delivery vectors are the new scarcity story.

The era of breakneck biologics CDMO capacity build-out in China is closing — not with a price cut, but with a shakeout. Recent industry reporting shows that six 15,000-litre reactors totalling 90,000 litres of capacity at Ledu Life Science Park in Shanghai's Baoshan district have been idle for over two years, and Ledu's biologics subsidiary was accepted into court-administered bankruptcy liquidation in December 2025.

For scale: Samsung Biologics' planned total capacity by 2032 is 1.324 million litres (up from today's 845,000), and Ledu's 2021 target once reached 1.3 million litres — a number Samsung still needs six more years to build, Ledu wrote into a deck five years ago. The logic then was simple: bigger reactors mean lower unit cost, orders will follow, and crossing the GMP and scale-up threshold is enough to win global outsourcing contracts.

The orders did not come. Foreign capacity exited even earlier: Lonza sold its Guangzhou biologics plant to Zoegene, and Boehringer Ingelheim let go of its 6,000-litre Shanghai site for no more than 20 million yuan — our earlier analysis put that secondary transaction at roughly 400 euros per litre, less than a third of replacement cost. Lonza's CEO put it plainly: Lonza does not compete in a capacity race.

Why the tank is no longer a moat

The idle capacity was not primitive: most of it held GMP licences, proven scale-up processes and cross-border compliance. But when capital replicated production lines from the same template, the moat degraded into an entry ticket — everyone had the tanks, the certificates and the ability to run conventional antibody tolling, leaving only price wars and heavy depreciation. Ledu's 90,000 stranded litres, the BI Zhangjiang site listed at a symbolic price, and Lonza's Guangzhou exit are three marks on the same liquidation curve: capacity that stops at commodified manufacturing no longer differentiates.

Nor is this uniquely Chinese. Samsung Biologics' binding CMO contract value grew by only about 200 million dollars in the third quarter even as project counts rose — dealers attribute the slow dollar growth to origin diligence under tariff and onshoring uncertainty. Lonza anchors its mid-term story on conjugation capability and margins rather than litres; Asymchem's management calls payload-linker capacity very tight. What is scarce has never been the reactor; it is process capability and delivery certainty in specific modalities.

Capital has not left; it has changed targets

On the same frozen ground, new money is being poured elsewhere. Within a year: Yinuo Bio closed two rounds totalling over 100 million yuan; PackGene has completed seven rounds with Sequoia China and with Sequoia China and SDIC Investment among the backers; Jinfan Medical took a nine-figure A+ round in April 2026. The shared keyword is not tonnage but vectors — AAV, LNP and other next-generation delivery technologies.

The split in logic is clean. Peptides and ADCs expand along proven demand, competing on scale, cost and delivery — a contest of volume after commercial validation. AAV and LNP players are positioning on next-generation delivery, betting on the scarcity of the capability itself: however elegant the mRNA sequence, without an LNP it never enters the cell; however clever the gene-therapy design, without a high-yield, low-empty-capsid AAV process it never becomes an accessible product. The centre of competition in gene therapy and mRNA is shifting from the molecule to delivery.

Three reminders for capacity holders

First, the dividing line in the shakeout is order certainty, not scale: the same 15,000-litre reactor carrying commercial ADC conjugation campaigns and one fielding conventional antibody enquiries are entirely different assets. Second, delivery capability is becoming the bottleneck asset for CGT and mRNA — and bottleneck assets are priced on irreplaceability, not depreciated by the litre. Third, for anyone still expanding, the three capacity-accounting questions we posed earlier still apply: which signed orders the new capacity maps to, whether ramp-period cash flow is self-funding, and what residual value the exit assumes — the 3-4% residual on BI's Zhangjiang site is the market's answer to the third.

The second half of China's biologics CDMO market is no longer a contest of who owns the biggest tank, but of whose capacity holds what no one else can put into it. Ledu's 90,000 litres and the vector startups' new rounds are two faces of the same coin.