Believer Meats raised $387 million, built what it called the world’s largest cultivated meat facility, and cleared both FDA and USDA. Its Wilson, North Carolina plant cost roughly $154 million. The baseline bid for that plant and its equipment is $50 million.

The number that matters more is not the discount. It is that the plant and the knowledge of how to run it are being sold by two different officials, in two different countries, under two different legal systems — and the Israeli trustee has said plainly that buying one does not get you the other.

We covered the failure itself in our post-mortem on the shutdown. This piece is about what the wreckage is worth, who is in line for it, and what the price implies for anyone still deciding whether to build capacity or buy it.

The two processes

US process Israeli process
Entity Future Meat Technologies Inc (FMI) Future Meat Technologies Ltd (FML), the parent
Official Receiver Kevin Sink Trustee Yoel Freilich, Gissin & Co
Court North Carolina Business Court Lod District Court
Started General receivership, order filed 6 February 2026 Insolvency petition filed 31 December 2025
Selling Plant, bioreactors, process tanks, centrifuges, freezing and wastewater systems, automation and production-control systems, permits and licences Three families of patents, cell lines, media formulations, scale-up know-how
Marketed by BDO Consulting Group Trustee, via data room and NDA
Bid deadline Extended from 20 July to 10 August 2026 Extended from 22 July to 5 August 2026
Next step Auction 17 August, sale hearing 20 August Not publicly scheduled

Freilich, on the separation: “whoever purchases the plant in North Carolina does not get the IP or the know-how of the company,” he told AgFunderNews, adding that a buyer does not have permission to use it unless it is bought from FML.

That is not a technicality. The flyer his office circulated describes what is for sale as “the technology and know-how in practical scale-up of cultivated-meat production: cell lines, media logic, bioreactor operation, perfusion strategy, centrifuge integration, automation recipes, harvest/wash procedures, and process data generated through Israeli pilot-scale development and transferred into large-facility design,” and argues that these assets “shorten the path from laboratory biology to a manufacturable platform.”

Strip that out and what remains in Wilson is a very large, very specific building with tanks in it.

The two officials do not agree

The US receiver’s position is the opposite of the trustee’s. In his brief supporting the sale motion, Sink wrote: “FML pleadings appear to imply that FML owns certain assets located at the facility, including certain intellectual property, although no such property was specifically identified. The receiver does not believe that FML owns any of the assets located at the facility and does not seek to sell any assets of FML.”

The February receivership order records the disagreement without resolving it — “The US receiver [Kevin Sink] does not agree, at this time, with all of the assertions of the trustee” — while preserving FML’s claims over excluded assets, a list that on AgFunderNews’s reading of the order expressly contemplates “any frozen cell lines located within the United States, media formulation receipts, and/or process diagrams and control software owned by FML.”

Read that carefully. The disputed assets are not abstractions held in Tel Aviv. Frozen cell lines physically inside the North Carolina building may belong to a different estate in a different jurisdiction. Sink also cannot sell any IP without Freilich’s written consent and the approval of the Lod District Court.

Practical consequence: a bidder pricing this facility is not buying a going concern at a discount. It is buying a shell plus an option, and the option requires winning a second auction, under Israeli insolvency law, whose bid deadline fell twelve days before the US auction is scheduled to open. Any valuation that treats the $50 million as the cost of acquiring Believer’s production capability is wrong by the price of the IP — a number nobody has published.

Who is in line for the $50 million

Two secured positions are named repeatedly in the filings.

Gray Construction designed and built the facility and alleges it is owed $36.4 million. When Believer fell behind, Gray accepted a $25 million secured promissory note backed by a mortgage on the plant, then agreed a forbearance deal with staged repayments beginning with $22 million due in December 2025. Believer defaulted on that immediately. Gray filed liens and pushed the company into receivership to reach its collateral.

Ameris Bank provided a $25 million term loan with a first-priority security interest over Believer’s bank accounts, machinery and equipment. It declared the loan due after the December default with more than $13 million outstanding.

Five further secured parties appear in the filings — Robert Reiser and Company, Leaf Capital Funding, ADM Ventures Investment Corp, GEA Systems North America and GEA Mechanical Equipment US — with no public figures attached. Separately, CSC Leasing has asserted ownership of selected machinery, which the court says the receiver cannot sell without CSC’s written authorisation. Roberts Oxygen kit is likewise excluded from the sale.

Our arithmetic, not the court’s: the two named senior positions come to roughly $38 million against a $50 million baseline. That leaves something in the region of $12 million before receivership costs, before the five unquantified secured parties, and before Gray’s unsecured balance of about $11 million. Equity — the $387 million — is behind all of it.

Practical consequence for investors: in a receivership of a capital-heavy food-tech business, the recovery question is settled by the construction contract and the equipment loan, not by the technology. Gray’s mortgage and Ameris’s blanket lien on machinery were both put in place after the money was raised, and both sit ahead of every venture dollar. When underwriting a company that is about to build a plant, the term sheet that determines your downside is the one with the contractor.

What the price says about building versus buying

The most useful reader of this auction is not a creditor. It is anyone currently modelling a new-build.

At the baseline bid, a completed plant with USDA and FDA inspections described by BDO as “complete and approved” is changing hands at roughly a third of what it cost to build. On capital raised, the ratio is 12.9%. Any greenfield capacity plan is now competing against that alternative.

UPSIDE Foods is the clearest illustration. It has raised $608 million. In early 2024 it paused its own large-scale plant in Glenview, Illinois, in favour of expanding its smaller Emeryville site, having conducted, on its own account, dozens of runs at 2,000-litre scale. AgFunderNews understands Believer’s bioreactors to be at the 20,000-litre scale. UPSIDE’s own framing is about speed: “This North Carolina facility would accelerate the expansion of our production capacity and help us meet growing demand for UPSIDE chicken much sooner.”

A company that declined to build 20,000-litre capacity at cost in 2024 is bidding for 20,000-litre capacity at roughly a third of cost in 2026, having demonstrated its process at one tenth of that volume. Whether the step from 2,000 to 20,000 litres works is the entire question, and it is not answered by owning the tank. It relates directly to the constraint we mapped in the fermentation capacity map: the gap is not bench capacity, it is the transition from a successful small run to a commercial line.

Practical consequence: distressed capacity resets the denominator for everyone. A capacity plan justified against build cost should be re-run against secondary-market cost, and the sensitivity that matters is not the purchase price but the cost and time of re-qualifying a plant designed around someone else’s process — with the process documentation in a different auction.

The numbers do not agree with each other

Three figures in this story are reported inconsistently, and the inconsistencies are instructive rather than trivial.

Figure Values reported Note
Capital raised $387m (insolvency filing, via Green Queen); “almost $400 million” (AgFunderNews) We use $387m, as the filing-derived figure
Plant cost $138m initial estimate, rising to ~$154m (insolvency filing); “Over $150 million invested in designing, building and installing related equipment” (BDO teaser) Different bases. Green Queen reads the ~$154m as excluding equipment; BDO’s figure explicitly includes equipment installation
Gray Construction claim $34m (first reported), $35.2m (Green Queen, January), $36.4m (court filings, February and July) Consistent with a claim accruing over time, but we have not seen the pleadings

The debt headline needs the same care. Coverage put Believer “nearly $225M in debt,” but the filing as reported breaks that into $11 million owed in Israel and roughly $213 million that the US subsidiary owed its own parent. An intercompany balance is not third-party debt. It matters for how proceeds are allocated between two estates that are already disagreeing, and it is not comparable to the $36.4 million Gray is owed.

This is the same problem we keep finding in this sector: figures that look additive are stated on different bases, and nobody says which. We ran into it with tonnage on unstated moisture bases and with cost per kilogram. Here it is in insolvency accounting.

The counter-argument

The case for reading this as a functioning market rather than a collapse is reasonable.

A stalking-horse bid is a floor, not a price. Its purpose is to guarantee a minimum while a banker tests the market, and the bid deadline was extended precisely because, in the receiver’s words, “Numerous discussions have occurred with interested parties and multiple parties have visited the facility.” A competing bid must clear $52.25 million, covering the purchase price plus breakup fees and expenses. The final number could be materially higher.

The court itself has said Believer’s “property has significant value that can be realized and applied to its various debts.” And Freilich’s stated reason for extending the IP deadline — “the vast interest shown in the technology from cultivated meat companies, investment funds, and other entities” — describes competition, not a fire sale.

The strongest version of the optimistic case: a purpose-built, regulator-inspected cultivated meat plant is being transferred to an operator with regulatory approval and $608 million raised, at a price that makes the unit economics far more survivable than they were for its builder. That is what a working secondary market is supposed to do. Our analysis of the supply-layer failures argued that capital quantity was not the differentiator; a lower cost basis is a real advantage, not a consolation.

What we could not establish

  • The final sale price. As of 12 August 2026 the auction had not taken place. Bids were due 10 August; the auction is scheduled for 17 August and the sale hearing for 20 August. We found no public report of how many bids were received by the 10 August deadline. Everything here describes a floor.
  • The outcome of the IP auction. Bids were due 5 August 2026. Re-checked on 12 August: we still found no public report of a winner, a price, or whether any bidder sought both the plant and the IP. Freilich declined to say whether facility bidders were also interested in the IP.
  • What the IP is worth. No guide price, reserve or valuation has been published for the patents, cell lines or know-how.
  • The full secured claim total. Five secured parties are named with no amounts. We cannot say what the aggregate claim on the proceeds is, so our headroom estimate is an upper bound on what is left after the two named positions, not a distribution forecast.
  • Whether $154m includes equipment. The insolvency filing figure is reported by Green Queen as excluding equipment; BDO’s “over $150 million” explicitly includes installed equipment. The two may describe the same spend on different bases or genuinely different spends. Our “roughly a third of build cost” ratio is sensitive to this and should be read as approximate.
  • The bioreactor scale. The 20,000-litre figure is AgFunderNews’s understanding, not a court document or a company statement. Court filings do not detail the bioreactors.
  • The IP ownership dispute. Two officers of two courts have stated opposite views about what FML owns inside the Wilson facility. No public ruling resolves it.

What to watch

  1. Whether the plant clears $52.25 million on 17 August. If it sells at the stalking-horse floor, that is the market’s answer on demand for large-scale cultivated meat capacity. If competing bids appear, the discount to build cost narrows and the buy-versus-build case weakens accordingly.
  2. Whether one party ends up owning both the plant and the IP. A split outcome creates a plant that cannot legally run its designed process and an IP holder with no facility. That would be the more consequential result, and the harder one to unwind.
  3. Whether the sale proceeds reach any unsecured creditor. On the figures published so far, we do not expect the $387 million of equity to recover anything, and we expect Gray’s unsecured balance to be impaired. Both are checkable once the receiver reports.

We will update this article with the auction outcome rather than writing a new one.