Ask anyone in fermentation-derived protein what the constraint is and you will get the same answer. GFI and Integration Consulting put it plainly in their 2023 capacity analysis: access to manufacturing capacity and the technical capabilities required at different development stages “has become a bottleneck for further growth.”
That diagnosis is three years old and nobody disputes it. So the more useful question is not whether the bottleneck exists. It is what happens when somebody tries to remove it.
Liberation Labs is the clearest test case available. It was founded in 2022 explicitly to build merchant precision-fermentation capacity in the United States, broke ground on a 600,000-litre plant in Richmond, Indiana in June 2023, and has assembled $126.5 million in private capital and non-dilutive commitments. Commercial production was set to begin by the end of 2024.
We could not confirm, in August 2026, that the plant is producing.
That is not an accusation. It is a description of the information environment, and it is the single most important thing an operator planning a scale-up needs to understand: the capacity everyone is counting on is being built, slowly, and its status is not public.
Methodology and limits
This piece assembles publicly disclosed capacity figures for fermentation-derived protein into one table. Three limits govern how far it can be read.
Every figure is a disclosure, not a measurement. None is audited. Companies announce nameplate capacity; utilisation is almost never published.
The figures are of different vintages, from August 2023 to March 2026. Each row carries its source date. A 2023 figure is not a 2026 fact.
The units are not commensurable. This is the finding that most constrains what the table can be used for, and we return to it below.
We have deliberately included only facilities where a specific number was published. There are more plants than these. A map of what is disclosed is not a map of what exists — but it is the map an operator, an investor or a journalist actually has.
The map
| Operator / facility | Role | Disclosed scale | Status as last publicly stated | Statement date |
|---|---|---|---|---|
| Liberation Labs — Richmond, Indiana | Merchant CDMO, precision fermentation | 150,000-litre fermenters, 600,000 litres total, with fully dedicated downstream processing; 600–1,200 tonnes protein/year projected | “Late stages” of construction; start-up targeted for 2025 | Feb 2025 |
| ENOUGH — Sas van Gent, Netherlands | Own plant, biomass fermentation | 200,000-litre tanks; one line at 10,000 tonnes/year, funded to double to 20,000; site ceiling stated as 60,000 tonnes | Producing since summer 2023; second line funded, ~18 months to install | Aug 2023 / Feb 2024 |
| Enifer — Finland | Own plant, biomass fermentation | Pilot line 5–10 kg/day, with contract-manufacturer campaigns at roughly 10× that; €33m plant for 3,000 tonnes/year | Commercial plant under construction; 2026 ramp-up approaching | Mar 2026 |
| iFAB Tech Hub — University of Illinois Urbana-Champaign | Public CDMO, pilot scale | Enough pilot capacity to test 60 new technologies per year | $51m Department of Commerce grant awarded 2024 | Jun 2025 |
| The Better Meat Co — Sacramento | Own demo plant, biomass fermentation | 9,000-litre bioreactor, 100+ harvests run | Demo scale; states it cannot commercially supply major CPG brands until it scales | Jun 2024 |
| Biovet AD (Huvepharma) — Bulgaria | Pharma-owned brownfield, rented to food | “Over 9,000,000L of installed fermentation capacity”; no food-grade sub-total or utilisation published | Producing EVERY’s OvoPro at “steady state” and “delivering metric tons to customers regularly”; agreement quadruples EVERY’s capacity | Jun 2026 |
Read the “disclosed scale” column and the problem is immediate. Liberation Labs is measured in litres of fermenter volume. ENOUGH and Enifer are measured in tonnes of output per year — on an unstated moisture basis. Enifer’s pilot is measured in kilograms per day. iFAB is measured in technologies tested per year, which is not a capacity unit at all.
You cannot add this column up. Nobody can. Any claim you read about total global fermentation capacity for alternative protein has resolved these units somehow, and the resolution is not published.
Practical consequence: when a supplier quotes capacity, ask for litres of working volume, annual output on a stated moisture basis, and current utilisation. Three numbers. If a supplier will give you only one, you have learned something about the other two.
What Liberation Labs actually built
The Richmond facility is worth detailing because it is the most thoroughly documented merchant fermentation asset in the category, and because the specification is genuinely well-suited to the problem.
It sits on a 36-acre parcel inside Phase I of the 700-acre Midwest Industrial Park in Richmond, close to Interstate 70, the Norfolk Southern mainline and three regional airports. The design centres on 150,000-litre fermenters totalling 600,000 litres of fermentation capacity, with a fully dedicated downstream process including separation and drying — which matters more than the fermenter number, because downstream capacity is routinely the binding constraint in food fermentation and is routinely omitted from capacity claims.
The BIO3 platform is engineered for merchant flexibility: multiple feed capabilities including sugar, methanol and glycerine; continuous sterilisation of media and carbon source; fed-batch and draw-and-fill operating scenarios; and the ability to handle genetically modified organisms. That last item is a food-CDMO differentiator, not a throwaway — a plant that cannot run GMOs cannot serve precision fermentation at all.
The economics as stated: $115 million capital cost, 600 to 1,200 tonnes of protein a year, $40 million in projected annual revenue, 45 direct manufacturing jobs each expected to support around three more in the region.
The financing and schedule record
Here is the sequence, as disclosed.
| Date | Event |
|---|---|
| December 2022 | $20m seed round led by Agronomics and Siddhi Capital |
| April 2023 | $30m equipment financing secured |
| June 2023 | Groundbreaking at Richmond |
| September 2023 | Front-end loading level 3 engineering complete, all major equipment ordered; commercial production expected late 2024 |
| October 2023 | $75m convertible note round announced |
| February 2025 | $50.5m closed — $31.5m new capital plus $19m of insider bridge notes from the prior year; construction described as “late stages”; start-up now targeted for 2025 |
Two things in that table matter more than the totals.
First, the $75 million round announced in October 2023 closed at $50.5 million in February 2025 — and Green Queen noted at the time that this came 15 months after the fundraising plans were first announced. Nineteen million dollars of the close was insider bridge notes raised the previous year, meaning a substantial fraction of the “new” round was existing investors keeping the project alive while the external raise took its course.
Second, the start-up target moved from “the end of 2024” to “2025” over roughly eighteen months, and we found no public statement after February 2025 confirming production began.
The full capital picture: $71.5 million in private capital since founding in 2022, plus $55 million in non-dilutive funding commitments — $30 million of equipment financing and a $25 million USDA loan guarantee. That is the $126.5 million figure in the headline, and it is a sum of the company’s own disclosed components rather than a figure the company has published in that form.
Practical consequence: if you are modelling a launch that depends on merchant US precision-fermentation capacity, do not treat announced nameplate capacity as available capacity. Ask for a commissioning certificate, a current utilisation figure, and the name of an existing customer running product through the line.
Who is renting, and from whom
GFI and Integration Consulting laid out the three routes to capacity: partner with a contract manufacturing organisation, build a greenfield site, or retrofit a brownfield facility. Their analysis is explicit that the third is underexploited, and specifically flags the opportunity to retrofit existing fermentation facilities in parallel industries to cut both the cost and the lead time of adding capacity. Breweries get their own section.
The reason that matters is arithmetic. Liberation Labs took three-plus years and $126.5 million to bring one 600,000-litre greenfield asset toward production. Retrofitting a plant that already has tanks, utilities, effluent handling and a trained workforce is a different order of problem — which is exactly why GFI devotes a figure to the CAPEX saved by brownfield development.
Update, 12 August 2026: the brownfield route produced the largest number in the table
The clearest example of the retrofit route GFI called underexploited arrived after this piece was first published, and it dwarfs everything above.
On 15 June 2026 The EVERY Company announced an expansion with Huvepharma, a pharmaceutical group, that quadruples EVERY’s production capacity for OvoPro, its precision-fermented ovalbumin. The production sits at Huvepharma’s subsidiary Biovet AD in Bulgaria, which EVERY’s release describes as “over 9,000,000L of installed fermentation capacity — one of the largest and most flexible in Europe.”
Set that against the rest of the table. Liberation Labs’ greenfield asset is 600,000 litres and took three-plus years and $126.5 million. Biovet’s installed base, as disclosed, is fifteen times larger and already exists, because it was built for animal and human health products rather than for food. Biovet’s chief executive director Anguel Jeliazkov put the proposition in one line: “We do not just scale processes — we industrialise them.”
Two caveats keep this from being a clean comparison, and both are the same problem this piece started with. First, 9,000,000 litres is the site’s total installed capacity, not the volume available to food customers; no food-grade sub-total and no utilisation figure is published. Second, EVERY’s demand-side number is stated without a base: it says that in the first four months of 2026 it secured annual orders worth 550% of its total 2025 order volume, and the 2025 volume is not disclosed. A ratio without a denominator cannot be compared to anything.
What is stated in absolute terms is modest and useful: EVERY says the collaboration is at “steady state production delivering metric tons to customers regularly.”
Practical consequence: if you are hunting for capacity, the shortest path in 2026 may be a pharmaceutical or animal-health fermentation site rather than a food-sector CDMO. Ask for the food-grade allocation and the utilisation, not the site total — the site total is the number that gets announced.
The renting tier is also where the public money has gone. In 2024 the US Department of Commerce funded the country’s first public CDMO with a $51 million grant to the Illinois Fermentation and Agriculture Biomanufacturing Tech Hub at the University of Illinois Urbana-Champaign, which is designed to provide enough pilot capacity to test 60 new technologies a year.
GFI’s own recommendation on how to spend that money is unusually specific and worth quoting for what it implies. Its guidance to policymakers reads: “Prioritize pilot, demonstration, and smaller commercial-scale bioreactors and processing equipment, given the number of companies operating at those stages and the general availability of bench-scale capacity.”
Translated: the shortage is not at the bench. It is in the gap between a successful bench run and a commercial line — the demonstration scale where a process either survives contact with real equipment or does not. That is precisely where The Better Meat Co sits with its 9,000-litre bioreactor, and precisely where it says it is stuck.
GFI also reports that the CDMOs it examined required $25 to $50 million in up-front public investment, with one reporting that it broke even after a single year of operation and earned back the full investment within several years. If that is generalisable, public pilot capacity is among the cheaper industrial interventions available. We have not seen the underlying case studies and cannot assess how representative that one CDMO is.
Practical consequence: a company at bench scale looking for its next step should be targeting public pilot CDMOs and brownfield retrofits, not waiting for merchant commercial-scale capacity. The commercial tier is where the delays are.
Three tiers, and what separates them
Sorting the map by position rather than by geography produces a cleaner picture.
Has tanks and is running them. ENOUGH, alone in this table. Its Sas van Gent line has been producing since summer 2023 at 10,000 tonnes a year. The reason it got there is covered in our analysis of what alternative-protein offtake commitments are actually worth: Cargill supplies its feedstock from the plant next door, takes its wastewater, holds a single-digit minority stake and resells the product. Co-location did more for ENOUGH’s capacity position than any financing round.
Building tanks. Liberation Labs, Enifer, iFAB. All funded, all under construction or recently granted, none confirmed at full commercial output in any source we found.
Stuck at demo. The Better Meat Co, which has run 100-plus harvests from a 9,000-litre bioreactor, has cut its at-scale costs by more than 30% through continuous processing, and still states it cannot commercially supply major CPG brands until it scales.
The variable separating tier one from tier three is not technology and not regulatory status. Better Meat’s process is arguably the most advanced in the table — continuous biomass fermentation, which its cofounder called “the holy grail for fermentation.” The variable is who is paying for the tank, and whether that party has a reason beyond return on capital to want it built.
The counter-argument
The strongest objection is that this piece measures the wrong thing.
Idle capacity is expensive, and a merchant CDMO that opens before it has committed volume burns cash on an empty plant. On that reading, Liberation Labs taking longer than announced is not a failure of execution; it is a company sequencing commissioning against a demand book that was itself slower to materialise than anyone forecast in 2022. The alternative — opening on schedule into a market that shrank — is how you get a distressed asset. Believer Meats built a $154 million plant and cleared regulatory approval before shutting down; we covered that in the post-mortem. Building slowly is not obviously the worse error.
A second objection: capacity announcements are marketing, and the absence of a “we have started production” press release is weak evidence of anything. Companies announce groundbreakings because they attract investors and local political goodwill. Commissioning a plant is unglamorous and often deliberately quiet while yields stabilise.
Both points are fair, and neither changes what an operator should do. If you are choosing where to make product in the next eighteen months, you need to know whether a line is running, and the public record does not tell you. The appropriate response to that is diligence, not inference in either direction.
What we could not establish
- Whether the Richmond facility is in commercial production. The last public statement we found is from February 2025, describing construction in “late stages” and targeting start-up in 2025. We found no confirmation of first commercial production, and no revised public timeline. We are not asserting the plant is not running — only that we could not confirm that it is.
- Current utilisation at any facility in the table. Not one operator publishes it.
- The moisture basis of the tonnage figures. ENOUGH’s 10,000 tonnes/year and Enifer’s 3,000 tonnes/year are stated without specifying dry or hydrated weight, so they cannot be compared with each other or with Liberation Labs’ 600–1,200 tonnes of protein.
- Whether ENOUGH’s second line was installed. In August 2023 CEO Jim Laird said it would take 18 months; in February 2024 the doubling was described as in progress. We found no confirmation of completion, so the 20,000-tonne figure should be read as funded rather than installed.
- Whether the $50.5m February 2025 round was the final close of the $75m convertible note round announced in October 2023, or an interim one.
- Whether Liberation Labs proceeded past the DoD feasibility stage. It received a $1.4m award under the Department of Defense’s Distributed Bioindustrial Manufacturing Program to study a four-million-litre flexible-use facility adjacent to Richmond, with up to $100m available at the build stage. We found no source confirming selection to build.
- How representative GFI’s break-even CDMO case is. One CDMO reportedly broke even in a year; we did not see the case studies and cannot judge whether that is typical or the best case in a small sample.
What to watch
- A confirmed first-production announcement from Richmond. It is the single most consequential piece of missing information in US fermentation infrastructure, and it either arrives or the delay becomes the story.
- Whether any operator publishes utilisation. Nameplate capacity without utilisation is a marketing number. The first company to publish both will be the first one a serious buyer can plan against.
- Whether brownfield retrofits start appearing in announcements. GFI has been pointing at breweries and parallel-industry fermentation assets since 2023. If the next wave of capacity is retrofit rather than greenfield, the lead times in this article stop being representative — which would be the best available news for the category.
This is a living reference. Where a figure here is superseded by a newer disclosure, we intend to update the table rather than write a fresh piece, and to note what changed. For the parallel question of what this capacity costs to run, see why four public cost models for precision-fermented protein span under $20/kg to $15,000/kg.