10. Technology fit: which platforms suit Vietnam

In one paragraph. No technology is a general answer for Vietnam. Each fits a specific combination of feedstock, buyer and rule. We score ten families against seven conditions drawn from the previous chapters. Two fit now: textured plant protein, because the rules are clear, the buyers exist and Vietnam imports what it would replace; and functional microbial ingredients for shrimp and fish feed, because Vietnam has the carbohydrate, the fermentation skills and a concentrated, motivated buyer base. Three are worth proving with pilots: local legume and rice protein ingredients, fungal foods such as koji and mycoprotein, and duckweed grown on aquaculture effluent for feed. Precision fermentation fits only for high-value, low-dose products made under contract, because public strains sit 5 to 15 times below the titer the literature expects for bulk proteins and the product probably counts as a GMO product in Vietnam. Cultivated meat and seafood remain research topics: Vietnam has no food cell line and costs are 7 to 15 times the retail price of meat. Insects are an established feed incumbent and appear here only as a benchmark.

---

#10.1 How we scored

Seven conditions, each rated strong, moderate or weak for Vietnam in 2026, on the evidence in chapters 2 to 9:

The ratings are our judgement VN-directMedium. The data behind each one are cited below and in the appendices.

#10.2 The fit matrix

Which technologies fit Vietnam

Ten families rated on seven conditions (strong, moderate, weak) with a verdict

#10.3 Plant protein

Textured plant protein. The technology is mature globally and cheap to install; the rules are ordinary food rules; and Vietnam imports what a domestic plant would replace. Every Vietnamese product with 10 g of protein or more per 100 g in our audit ran on imported soy protein, textured soy or wheat gluten, mostly from China ^1,2,3,4 VN-directHigh. We found no commercial textured-protein extrusion in Vietnam ^5,6,7 VN-directMedium and at most one Vietnamese extrusion study ^8,9 VN-directMedium. Buyers are named: frozen-food makers and vegetarian (chay) brands such as An Nhien, Au Lac, Xuan Hong, LC Foods, Cholimex, CJ Foods Vietnam, Vissan and Thoai An, plus co-packers ¹ VN-directMedium. The weaknesses are feedstock (soy is about 98% imported, and we found no food-grade line at the large Phu My crusher ¹⁰ VN-directMedium) and competition from Chinese textured protein at about USD 1.6 per kg ² VN-directMedium. High-moisture extrusion of whole-cut analogues is a second step once a dry line and a customer base exist.

Local legume and rice proteins. Mung bean is the main protein in mass-market chay fillings, and Vietnam imported USD 138 million of dried mung and black gram beans in 2023, 65% of it from Myanmar ¹¹ VN-directMedium. Protein quality is reasonable in pigs (DIAAS: soy isolate 97, mung bean 86, pea concentrate 70, brown rice concentrate 42) but human digestibility of mung bean measured in children was about 20% lower than the pig value ^12,13,14 generalHigh. No measured rice-plus-legume blend reaches a DIAAS of 100 ¹⁵ generalMedium. The plausible first market is replacing imported pea protein in high-protein plant drinks and yogurts, not replacing textured soy ¹ VN-directMedium. Competition from Chinese pea and rice proteins is strong, so this needs a buyer specification and a pilot before a plant.

#10.4 Microbial protein for feed

The science supports partial, not full, replacement. In whiteleg shrimp, gas-fermentation protein replaces about 30% of fishmeal without growth loss; halving fishmeal cut growth unless a bile acid was added. Methanotroph meal had lower digestibility in one Chinese screen ^16,17 generalHigh. Black soldier fly meal did no better (growth fell at 30% replacement) ¹⁸ generalHigh. A 2025 preprint reports mixed-culture protein grown on soybean-processing wastewater replacing up to 90% of fishmeal in shrimp, directly relevant to Vietnam's tofu and soy-milk industry but not yet peer reviewed ¹⁹ VN-adjacentMedium.

Function sells better than protein. Many benefits reported for novel proteins in shrimp and catfish are functional: better survival after pathogen challenge, palatability, or winter survival with beta-glucan (78% to 95% in striped catfish) ^20,21,22,23 generalMedium. Economics shows why that matters: bulk fishmeal replacement is marginal on price, while an ingredient sold at 1 to 5% of the diet for survival or feed efficiency can command a large premium: a pangasius mill could pay about USD 1,000 to 4,250 per tonne in total, and a shrimp mill USD 1,280 to 4,800 per tonne above fishmeal parity.

Fit in Vietnam. Carbohydrate is abundant (Asset map); aerobic fermentation skills and tanks exist inside amino-acid producers (Industrial base); about six firms make most shrimp feed (Industrial base); and inactive yeasts are already on the livestock feed list, while bacterial single-cell protein and fungal biomass need a feed trial and the aquafeed list has a drafting gap (Rules) ^24,25,26 VN-directHigh. Thermotolerant yeasts such as Kluyveromyces marxianus, isolated from Thai cassava pulp, grow at 40 °C on raw cassava starch ^27,28 VN-adjacentHigh, which cuts the cooling penalty. The weakness is competition: Chinese by-product bacterial biomass is already sold in Vietnam ²⁹ VN-directMedium.

Gas fermentation (hydrogen and CO2, methane, industrial off-gas) is commercial in China for feed ^30,31 generalMedium, but Vietnam has no measured supply of suitable gas and a carbon-heavy grid (Economics). Watch.

#10.5 Fungal foods

• Koji and solid-state fermentation. Vietnam has a tương (fermented soybean paste) tradition, research groups with gene tools for Aspergillus oryzae ^32,33 VN-directMedium, and a clear route: koji on food substrates goes through ordinary self-declaration (Rules). A. oryzae cannot make aflatoxin but can make cyclopiazonic acid and 3-nitropropionic acid; industrial strains carry a deletion that removes the risk, so a koji business needs a characterised strain, not a wild starter ³⁴ generalHigh.
• Mycoprotein and yeast food protein. Published costs are about USD 29.6 per kg of protein ³⁵ generalHigh, and every fungal or yeast food protein carries an RNA-reduction step (about 10% to 2% of dry weight by a 68 °C heat step) ³⁶ generalHigh. Published yields need care: the widely repeated "118.5 g/L" A. oryzae biomass on vinasse is not a broth concentration; the plausible flask value is about 5.9 g/L ³⁷ generalHigh. Chinese plants at 11,000 to 20,000 t a year set the price (Region). There is no novel-food route for a new fungal species (Rules).

#10.6 Duckweed and microalgae

• Duckweed (bèo). Protein is 20 to 30% of dry weight in Wolffia grown in the laboratory, up to 44% in Thai Wolffia ^38,39 generalHigh. Planning yields are 10 to 30 t of dry matter per hectare per year ^40,41 generalMedium. Duckweed is already a permitted livestock feed material ^24,25 VN-directHigh, and Vietnamese researchers hold more than 100 wild samples ⁴² VN-directMedium. For food, manganese is the gate: EFSA could not establish the safety of whole Wolffia powder in 2021 because of manganese intake, but judged a Lemna protein concentrate safe in 2023 ^43,44,45 generalHigh. A "PDCAAS 0.89" figure repeated in earlier drafts has no traceable source ³⁸ generalLow. Fit: feed on aquaculture effluent now (with heavy-metal testing), food as a protein concentrate later.
• Microalgae. Vietnam has small spirulina producers selling supplements, a salt-tolerant spirulina strain reaching 48.7% protein, and a research base at regional level ^46,47,48 VN-directMedium. Algae are on the feed list and spirulina is an ordinary food. But production is small and costly, and one source claims most spirulina sold in Vietnam is imported ⁴⁷ VN-directLow. A niche with incumbents, not a bulk protein route.

#10.7 Precision fermentation

• Titers. The leading review says bulk food proteins need titers "above 50 g/L", an expert assertion rather than a modelled threshold ^49,50 generalHigh. The best published leghemoglobin titers are 3.5, 7.27 and 10.1 g/L ^51,52,53 generalHigh, 5 to 15 times lower. Heme proteins are used at 0.5 to 2% of a product, so the benchmark binds less for them.
• Strains. OPENPichia offers a Pichia host under a liberal licence allowing royalty-free commercial manufacture; users still accept a licence, and product patents still apply ⁵⁴ generalHigh.
• Rules. Precision-fermented proteins probably count as "products of a GMO" under Decree 43/2026, and production with a GM strain needs a closed-production certificate whose timeline is unpublished (Rules).
• Infrastructure. Vietnam has no food-grade contract fermentation and no open pilot fermenter (Industrial base); the nearest are in Singapore, Thailand and China (Region).

Fit. High-value, low-dose ingredients (enzymes, heme and sweet proteins, specialty functional proteins), developed by Vietnamese groups that already express proteins in yeast (Knowledge and talent) and made under contract abroad until a domestic tolling option exists. Not commodity dairy or egg proteins.

#10.8 Cultivated meat and seafood

• No food cell line. The striped catfish lines that exist (India, Taiwan) were built for virology and toxicology. The best documented, PHT, needs 20% fetal bovine serum and grows attached to surfaces ⁵⁵ VN-adjacentHigh; the others are known only from their published descriptions ^56,57,58 VN-adjacentMedium. The "shrimp" line PmLyO-Sf9 is a black tiger shrimp and insect hybrid ^59,60 generalHigh. No true continuous shrimp cell line was found.
• Cost. Modelled cost is USD 37 to 51 per kg of wet cells ⁶¹ generalHigh, 7 to 15 times the retail price of lean pork or chicken breast in Vietnam (Economics). The best documented animal-free culture reached about 28 million cells per mL, below the densities cost models assume ^62,63 generalMedium.
• Rules. No route in Vietnam; the nearest approvals are in Singapore (Region).

Fit. Research, not production: serum-free adaptation of striped catfish cells, growth-factor expression (which Vietnamese groups already do), and regulatory preparation. Watch Singapore and Korea for signals.

#10.9 Insects: the benchmark

Black soldier fly meal is the one alternative protein made at industrial scale in Vietnam, with an 11,000 t a year design capacity and a seafood exporter's offtake (Alt protein today, Capital). It is on the livestock feed list. As agreed in scope, this report treats insects as the incumbent that every microbial or plant feed protein must match on price, digestibility and supply security, not as a play we recommend.

#10.10 What this means

• Investors: the two "build now" families have modest capital needs and B2B buyers. Everything else needs a pilot, a partner or a rule change first.
• Startups: pick a family where the route to market is already clear (textured protein, listed feed materials, koji) and sell function or import substitution, not novelty.
• Manufacturers: textured-protein supply and functional feed additives are the two families where a Vietnamese supplier could reach you within two years.
• Research bodies: the most useful work sits in the "prove first" row: DIAAS for local proteins, duckweed manganese and heavy metals on Mekong water, soy-wastewater protein in shrimp, thermotolerant strains on cassava streams, and characterised tương strains.
• Policy makers: three families (fungal foods, precision fermentation, duckweed food) move from "weak" to "moderate" on route to market with one change: a proportionate new-food procedure (Policy options).

Related: J. Science (the evidence family by family), Plays (plays built on this matrix).