The ranking

In one paragraph. Six categories were assessed against the same seven questions: is the raw material available in Vietnam, is there a legal route to market, how much capital does it need, does the country have the infrastructure, is the technology ready, does it transfer to a hot lower-middle-income country, and how good is the evidence. The order below is the answer. It is a ranking of what Vietnam should work on, not a ranking of which technologies are most exciting.

1. Fungal or yeast biomass protein on cassava processing residue, sold as feed first#

The only category where raw material, process maturity, cost structure and existing industrial capability line up at once. The feedstock sits at roughly 120 identified starch factories. The process is enzymatic starch hydrolysis, which those factories already run. Published cost models for bulk microbial protein reach USD 1.20 to 2.07 per kilogram at scale, though the lowest of those figures comes from a microbial oil process and the highest from gas fermentation, so neither is a like-for-like model of this route. Vietnamese manufacturing labour is roughly a tenth, and Vietnamese electricity roughly 40 percent, of what those models assume, though chapter 10 explains why the second of those is worth less than it looks.

The domestic framing is much stronger than the export one: Vietnam imported 21.75 million tonnes of feed grain in 2024, produced 21.5 million tonnes of compound feed, and imports essentially all of its soybean meal protein.

Blocked by: no novel food pathway, an undocumented approval timeline even on the feed side, no pilot facility in the country, no Vietnamese techno-economic assessment, and a nitrogen input that has to be bought in because cassava pulp has almost none.

2. Rice protein, entering through broken rice rather than bran#

Second because the raw material is uniquely Vietnamese at scale, is concentrated at mills rather than scattered across fields, and is hypoallergenic, which nothing else in the plant-based set is.

The entry point is broken rice and milled endosperm, which reach 78 to 85 percent protein and fit existing starch-processing lines. Rice bran is the more interesting research problem and the weaker commercial proposition: it needs defatting and heat stabilisation, and no published extraction route achieves yield and purity together. See chapter 11.

This is a research bet rather than a build bet, and on the bran side it is the one place where a Vietnamese programme would plausibly be doing something the field needs rather than importing a solved process. Vietnamese broken rice volumes and prices are unestablished: open question NP-G7.

3. Precision fermentation of heme or dairy proteins, entered through a regional contract manufacturer#

The science is far from mature. Published titers run from 35 milligrams to 7.27 grams per litre against a stated economic threshold above 50 grams per litre, and precision-fermented milk protein costs USD 210 to 310 per kilogram against USD 15 to 25 for the conventional article.

Three things nonetheless put it third rather than last. OPENPichia removed the licensing barrier to the standard production yeast. The Food Industries Research Institute already runs that yeast for enzyme expression. And ScaleUp Bio in Singapore operates a licensed food-grade precision fermentation facility with up to 10,000 litres of capacity that already takes third-party clients from Australia, the United States and Malaysia.

The way in is a contract, not a construction project.

4. Novel plant-based protein from mung bean, sesame, peanut, coconut or cashew#

Good protein science, no local crop at scale. Vietnam crushes 34,000 tonnes of peanut and 24,000 tonnes of copra a year, and mung bean falls below the USDA reporting threshold. Cashew is the exception worth checking, because Vietnam has the world's largest cashew processing industry, though the tonnage of broken kernels is unpublished and cashew is a major allergen.

5. Gas fermentation on hydrogen-oxidizing bacteria#

Carries the lowest minimum selling price of any dedicated protein process found in this study, USD 2.07 per kilogram, and has a product on sale in Singapore since 2022. Two lower figures appear elsewhere in this report, at USD 1.20 and USD 0.81 per kilogram, but both come from a microbial oil model rather than a protein one, so they are not directly comparable.

It ranks low for Vietnam specifically because it needs electrolytic hydrogen, pressurised gas systems and a supply chain the country does not have, and its economics rest entirely on cheap renewable electricity.

6. Cultivated meat. Watch, keep the regulatory door open, do not build#

Achieved cell densities are 10^6 to about 10^7 cells per millilitre, with specific reports of 2 to 3 × 10^6 for muscle cells on microcarriers and about 2 × 10^7 in small stirred flasks, against cost models that assume above 10^8. Demonstrated bioreactors run from tens to low thousands of litres while models assume 20,000 litres or more. Growth factors are commonly 90 to 95 percent of the cost of the culture medium.

The bottlenecks are linked: cheap medium, high cell density and large vessels all have to improve together, and no demonstrated system shows more than part of that package. For a country with no existing players, entering now means funding frontier science without any of the inputs that would make it cheaper to do in Vietnam than elsewhere.

On the two framings#

The domestic framing is where the numbers work, and it is a feed story before it is a food story. The export framing works only for precision-fermented ingredients routed through Singapore, and only because Vietnam cannot currently sell a novel food domestically at all. That asymmetry is itself an argument for running the regulatory work in parallel with anything technical.