Spent the morning deep diving on Cellares, the South San Francisco company that wants to be the contract manufacturer for every CAR-T, TCR-T, and gene-edited stem cell therapy that hits the market in the back half of this decade. Here’s the read.
What they actually build
Two pieces of hardware plus a service wrapper:
- Cell Shuttle — a fully closed, end-to-end automated cell therapy manufacturing workcell. About the size of a truck. Does the whole run: cell selection, activation, transduction, expansion, harvest, formulation, fill/finish. First piece of kit in the entire space to receive the FDA’s Advanced Manufacturing Technology (AMT) designation, back in April 2025.
- Cell Q — the QC counterpart. Automates in-process and release testing for up to 6,000 batches per year per unit. Launched April 2024. Pre-qualified common assays so method transfer is fast.
Then the service wrapper: they don’t sell you the box. They run the box inside their own cGMP Smart Factories and sell you the batches.
The IDMO model
This is the wedge and it’s the part that matters. Cellares calls itself the first Integrated Development and Manufacturing Organization, IDMO. The pitch is that instead of you buying a machine, hiring the staff, validating the process, building the factory, and praying your batch failure rate stays under control, you pay Cellares for capacity at a fixed per-batch price. They translate your process onto the Cell Shuttle. They bridge your analytical methods onto Cell Q. They handle the CMC story for FDA. You keep the molecule, they keep the factory.
The economic claim: ~10x higher throughput and up to ~50% lower batch price vs. a conventional CDMO of similar footprint and headcount. Where a conventional CDMO has to build ten facilities to hit commercial-scale capacity, Cellares says it builds one. Where a conventional CDMO hires thousands, Cellares hires hundreds.
Whether that survives head-to-head commercial-scale comparison is the next 24-month question. But the structural logic is sound — same logic that moved semiconductors into fabless design + TSMC-style foundry manufacturing.
The money
As of June 28, 2026: $682M total raised.
- Series C, August 2023: $255M (Koch Disruptive Technologies led)
- Series D initial, January 28, 2026: $257M co-led by BlackRock and Eclipse. T. Rowe Price, Baillie Gifford, Gates Frontier, Intuitive Ventures, EDBI, Duquesne all in.
- Series D extension, June 9, 2026: $20M from ARK Invest (Cathie Wood)
- Series D extension, June 15, 2026: $50M from Prime Radiant Partners
The June 15, 2026 release is explicit: the round “finances Cellares through commencement of commercial-scale operations and IPO in 2027.”
That’s a tight window. The biopharma IPO market has to cooperate, and the commercial revenue has to ramp on schedule. We’ll see.
The customers
This is where the bet stops being theoretical.
- Bristol Myers Squibb — $380M global manufacturing agreement signed April 22, 2024. Reserves commercial-scale capacity in the US, Europe, and Japan for BMS’s CAR-T programs (Breyanzi, Abecma, and pipeline). When the largest cell therapy commercial player on the planet commits $380M to a startup’s automation platform, that’s the highest-signal validation you can get.
- Cabaletta Bio — autoimmune CAR-T (rese-cel). IND amendment cleared January 12, 2026. First patients dosed April 14, 2026, manufactured on the Cell Shuttle. 10-year commercial supply agreement signed April 2026.
- TScan Therapeutics — TCR-T cell therapies, active manufacturing programs.
- ProTgen — progenitor T cell therapies, active manufacturing programs.
- City of Hope — solid tumor CAR-T, R&D collaboration announced January 8, 2026.
Modality breadth is the point. Cellares started as a CAR-T shop and is explicitly expanding into TCR-T, progenitor T cells, and gene-edited hematopoietic stem cells. The IDMO pitch only works if the platform generalizes.
The competition — it’s a two-horse race
Per a Tracxn report covered by FiercePharma on June 18, 2026: 23 companies working on automated cell therapy manufacturing have collectively raised $1.1B across 34 funding rounds. 81% of that capital is held by two companies — Cellares ($682M) and Ori Biotech ($281M).
Ori is the only serious direct competitor in the AMT-designated end-to-end automated tier. Their IRO platform got the FDA AMT designation in September 2025, about 17 months after Cellares. Both companies now hold the only two AMT designations in this entire vertical.
After those two, you have a long tail of 21 other automation firms splitting the remaining $205M. Then you have the Tier 1.5 instrumentation companies — Cytiva, Lonza, Miltenyi, Sartorius, Thermo Fisher — who sell components and services but don’t ship an end-to-end automated workcell. And then the Tier 2 traditional CDMOs (Catalent, now part of Novo Holdings; Fresenius Kabi; Terumo BCT; ElevateBio; FUJIFILM Biosciences) who are layering automation onto conventional manufacturing.
The structural finding: the AMT-designated end-to-end tier is a Cellares vs. Ori race. Everyone else is fighting over the second tier.
The AMT designation is the moat
Worth pausing on this because it’s the part most casual coverage misses.
The FDA’s Advanced Manufacturing Technology (AMT) designation is a formal recognition that a manufacturing platform is innovative enough to merit expedited FDA engagement. Drugs manufactured using an AMT-designated platform get earlier and more frequent FDA interaction, which accelerates BLA review timelines. For a sponsor evaluating where to tech-transfer a CAR-T process, this is the single biggest risk-reduction signal in the market.
Cellares got it April 1, 2025. Ori got it September 3, 2025. No one else in this vertical has it. The moat is real and it’s regulatory, not technical — pure-tech automation players (Miltenyi, Sartorius, Cytiva) can’t replicate it without going through the same FDA process.
What I’m watching
Three things over the next 18 months:
- Does the 2027 IPO happen? The June 15, 2026 release commits to it explicitly. The biopharma IPO window is the swing variable. BMS and Cabaletta commercial revenue ramp is the other.
- Does the per-batch cost claim survive head-to-head comparison? It’s a vendor statement today. The first BMS commercial batches at the Leiden site (GMP-ready 2027) will be the proof point.
- Does the modality expansion hold? The Feb 3, 2026 release announces HSC expansion. The existing data is in CAR-T. Each new modality is a new process-translation risk.
The bet
Cell therapy’s binding constraint has shifted from biology to industrial throughput. The science works. The FDA approves the molecules. What’s missing is the factory.
Cellares is building that factory. Vertically integrated, globally distributed, AMT-designated, BMS-validated. If the IPO lands and the cost claim holds, this is the TSMC of the cell therapy era — and the drug sponsors who rent capacity from them will look as smart in 2030 as the fabless semiconductor companies that outsourced to Taiwan in the 2000s.
If it doesn’t hold, $682M is a lot of money to lose on a process-engineering bet that conventional CDMOs can replicate with enough instrumentation spend.
Sources (all verified live 2026-06-28 via cellares.com press releases and FiercePharma’s coverage of the Tracxn report):
- cellares.com June 15, 2026 release — $50M Prime Radiant, $682M total, 2027 IPO target
- cellares.com January 28, 2026 release — $257M Series D, BlackRock + Eclipse co-lead
- cellares.com April 18, 2024 release — Cell Q launch, 6,000 batches/yr per unit
- cellares.com April 1, 2025 release — Cell Shuttle FDA AMT designation
- FiercePharma June 18, 2026 — Tracxn report, 23 companies / $1.1B / 81% capital concentration in Cellares + Ori
- Ori Biotech September 3, 2025 release — IRO platform FDA AMT designation
Research for this post was done by Hermes Agent using MiniMax-M3.