Best Probiotics with Akkermansia (2026 Buyer's Guide)

We compared the Akkermansia muciniphila supplements actually sold in the US on five criteria: dose against the one human trial, live versus pasteurized cells, whether the count is guaranteed through expiration, testing you can see, and real cost per day. Pendulum has the most human-trial history. For most people who want Akkermansia plus a conventional probiotic in one capsule, the value pick is ours, and the math is on the page so you do not have to take our word for it.

Mt. Angel Vitamins makes one of the products on this page. We ranked it with the same five criteria we applied to everyone else, we link the lab page that holds its lot testing, and where another brand is the better choice for a situation, that brand gets the category. Pendulum takes the clinical-validation pick below, and we say why.

If you only read one thing

For most people who want Akkermansia and do not already take a probiotic: a pasteurized (non-live) Akkermansia product with a disclosed count, third-party testing, and a cost under $1.50 a day. Our 3-in-1 Probiotic Fusion covers that in one capsule at $1.40 a day.

If you want the live bacterium from the company with the most human-trial history: Pendulum Akkermansia, at $1.80 a day.

If you only want pasteurized Akkermansia at the lowest cost: Toniiq Akkermansia 50 at $0.83 a day.

Scroll for the situation picks and the side-by-side.

Key takeaways

• In the only randomized human trial of Akkermansia muciniphila on its own (32 completers, three months, Depommier et al., Nature Medicine, 2019), the pasteurized, non-living cells produced the significant results and the live cells did not.
• The human-studied dose is 10 billion cells per day. The five products here range from 100 million live cells to 50 billion pasteurized cells per serving, a 500-fold spread.
• Akkermansia counts are not all the same unit. CFU counts cells that grow on a plate, AFU counts live cells by flow cytometry, TFU counts every cell, living or not. Compare like with like or you are comparing nothing.
• Only two of the five products print a strain code on the label (Metagenics: MucT; Double Wood: AH39). None of the five publishes a per-lot Akkermansia count at end of shelf life on its product page.
• Cost per day runs from $0.83 (Toniiq, Double Wood) to $2.33 (Metagenics). Price per bottle tells you nothing until you divide by servings.
• Pasteurized Akkermansia muciniphila has a 90-day rat toxicology study behind it and a 2022 EU novel-food authorization capped at 34 billion cells per day for adults, excluding pregnant and breastfeeding women. One product here is above that ceiling.

How we judged them

Five criteria, applied the same way to every product including ours. Each one has a number attached, because a criterion without a number is an opinion.

Dose against the human trial

The only randomized, placebo-controlled human trial of Akkermansia muciniphila by itself used 10 billion cells a day for three months. Forty overweight adults with insulin resistance enrolled and 32 finished (Depommier et al., Nature Medicine, 2019). Compared with placebo, the pasteurized arm improved insulin sensitivity by about 29 percent, cut fasting insulin by about 34 percent, and lowered total cholesterol by about 9 percent. Body weight fell 2.3 kg against placebo but did not reach statistical significance (p = 0.09). The live arm did not produce the same significant changes. The limitation that matters: one center, 32 people, three months, and a population that already had insulin resistance, so the result says nothing about healthy adults.

That trial sets the benchmark at 10 billion pasteurized cells per day. A product at 5 billion is at half the studied dose. A product at 100 million is at one percent of it. A product at 50 billion is five times it, which is not automatically better, and we come back to that under safety.

The mouse work that started all this points the same direction with one twist. The 2013 study that first showed a causal effect found that heat-killed cells did nothing in mice; only live cells improved the metabolic profile and thickened the mucus layer (Everard et al., PNAS, 2013). Three years later the same group found that pasteurizing the bacterium, a gentler treatment than the heat-killing used earlier, made it work better rather than worse in mice, and traced part of the effect to a heat-stable outer-membrane protein called Amuc_1100 (Plovier et al., Nature Medicine, 2017). "Dead" is not one thing. How the cells were killed matters.

Live versus pasteurized, and what the count unit means

Akkermansia muciniphila was isolated in 2004 from human stool as a strictly anaerobic bacterium that eats mucin, the protein in your gut's mucus layer, which means it lives on the lining of your intestine and dies in air (Derrien et al., Int J Syst Evol Microbiol, 2004). It is not rare. It makes up more than 1 percent of all the bacterial cells in a typical adult's stool (Derrien et al., Appl Environ Microbiol, 2008), colonizes infants within the first year of life at close to adult levels, and declines in the elderly (Collado et al., Appl Environ Microbiol, 2007). Later work showed it tolerates and even uses oxygen at nanomolar concentrations, the thin band that leaks from your gut wall, not the air inside a capsule (Ouwerkerk et al., Appl Environ Microbiol, 2016). Keeping it alive in a bottle is a real engineering problem. The published lab protocol for storing viable cells involves glycerol and minus 80 °C (Ouwerkerk et al., Beneficial Microbes, 2017).

That is why the count units differ, and why you have to read them:

• CFU (colony-forming units) counts cells that grow into colonies on a plate. The classic probiotic number. Hard to use for a strict anaerobe that grows poorly outside the gut.
• AFU (active fluorescent units) counts cells with intact membranes by flow cytometry. It captures live cells that would never grow on a plate. Pendulum uses this.
• TFU (total fluorescent units) counts every cell the cytometer sees, living or dead. It is the right unit for a pasteurized product, because the cells are intentionally not alive, and the wrong unit for proving anything is alive.

A label that says "5 billion TFU" and a label that says "100 million AFU" are not describing the same thing. A brand that prints TFU while claiming the cells are alive is mixing its units.

Count guaranteed through expiration

Bacterial counts decay. A label number that is true at the factory and false at the back of the shelf is not useful. We looked for an explicit statement that the count is guaranteed through expiration and for whether the brand publishes testing to show it. Metagenics states it. Pendulum says live strains are "verified at intended levels, through shelf life." Our own lab page shows lot testing through shelf life. The others either do not say, or say it only in marketing copy with no document behind it.

Testing you can actually see

Three tiers. The best: a certificate of analysis (COA) you can download, per lot, showing identity, count, and contaminant results. The middle: a named testing program with a document somewhere on the site. The bottom: a "third-party tested" badge and nothing else. If you have never read a COA, our guide to supplement testing and label claims walks through what each line means, and how we qualify a raw material before it goes into a capsule shows what the process looks like from the manufacturer's side. Where a brand links a COA we say so. Where it does not, we say that too.

Cost per effective day

Price per bottle divided by servings per bottle, using today's one-time price on the brand's own site (verified October 14, 2026), not the subscription price. Then we look at what that day buys: how many cells, in which unit, of which strain, and whether anything else rides along in the capsule. A $25 bottle that lasts 30 days costs the same as a $50 bottle that lasts 60.

Safety, briefly

Pasteurized Akkermansia muciniphila has more formal safety work behind it than most ingredients on the supplement shelf. A 90-day rat study at up to 1,500 mg/kg/day, about 96 billion cells per kilogram of body weight, found no adverse effects at any dose, and two genotoxicity assays were negative (Druart et al., Journal of Applied Toxicology, 2021). On that evidence the European Food Safety Authority concluded the ingredient is safe for adults at up to 34 billion cells per day, excluding pregnant and breastfeeding women, and the EU authorized it as a novel food on 1 March 2022 (Commission Implementing Regulation (EU) 2022/168). That is a European ceiling and the FDA has no equivalent number, but it is the only regulator-reviewed limit that exists, and it is the reason a 50 billion TFU serving gets a flag below.

Live Akkermansia has less formal safety work, though the Depommier trial reported both the live and pasteurized arms as safe and well tolerated over three months.

Evidence at a glance

Mechanism notes, for the researcher. The bacterium adheres to human gut cells in culture and raises the electrical resistance of the cell layer, a standard lab measure of barrier tightness (Reunanen et al., Appl Environ Microbiol, 2015). Its outer-membrane protein Amuc_1100 signals through Toll-like receptor 2 and raises the anti-inflammatory cytokine IL-10 in cell models (Ottman et al., PLoS One, 2017). A separate secreted protein, P9, triggers GLP-1 release from gut cells and in mice (Yoon et al., Nature Microbiology, 2021). In 49 overweight adults on a calorie-restricted diet, people who started with more Akkermansia in their stool had the healthier metabolic profile and improved more (Dao et al., Gut, 2016), which is an association, not a supplement result. Two reviews from the group that discovered the organism cover the full arc from isolation to the human trial (Cani and de Vos, Frontiers in Microbiology, 2017; Cani et al., Nature Reviews Gastroenterology & Hepatology, 2022). All of that is mechanism and association, not outcome. None of it is a product claim.

The picks

Best All-in-One for Most People: Mt. Angel Vitamins 3-in-1 Probiotic Fusion

Akkermansia: 5 billion TFU per capsule, pasteurized (the label reads "Akkermansia muciniphila Postbiotic," and the product page calls it a standardized non-live preparation). No strain code printed. Also in the capsule: 13 live Lactobacillus and Bifidobacterium strains totaling about 20 billion CFU (L. rhamnosus LRa-05 at 5 billion and L. reuteri LR08 at 4 billion are the two largest), 100 mg FOS prebiotic, 25 mg champignon mushroom extract. Form: one vegetarian capsule daily, shelf-stable, no refrigeration. Servings: 30 ($41.99) or 60 ($71.99). Cost per day: $1.40, or $1.20 on the 60-count. Verified October 14, 2026.

This is the only product in the group that puts pasteurized Akkermansia and a full multi-strain probiotic in one capsule: the postbiotic supports a healthy gut barrier while the live strains support digestive balance.* If you were going to buy a probiotic anyway, the Akkermansia comes along without a second bottle. How Akkermansia fits alongside a conventional probiotic has its own post, and our head-to-head with Seed DS-01 covers how it compares with a synbiotic that has no Akkermansia at all.

Made in our own cGMP facility in Mt. Angel, Oregon. Every batch is tested in our in-house lab for identity, potency, and purity, and certificates of analysis are published on our lab testing page.

Where it falls short: 5 billion TFU is half the human-trial dose, and the label does not print a strain designation. If Akkermansia is the only thing you want, Toniiq gives you ten times the pasteurized count for less money.

Mt. Angel Vitamins 3-in-1 Probiotic Fusion, $41.99 for 30 capsules

Best Clinically Validated Live Akkermansia: Pendulum Akkermansia

Akkermansia: 100 million AFU per capsule, live, single strain. No strain code printed on the product page. Also in the capsule: nothing else active. Form: one capsule daily; the brand says "refrigeration is best though not required." Servings: 30 ($54.00, shown against a $75 compare-at price), 90 ($145.00), 180 ($275.40). Cost per day: $1.80 on the 30-count, $1.61 on the 90-count. Testing: every batch tested by independent accredited labs; live counts verified at intended levels through shelf life; Non-GMO Project Verified. Verified October 14, 2026.

Pendulum wins this category because it is the only company on the list that has run a randomized, placebo-controlled human trial of a formulation containing its own Akkermansia strain: 76 adults, 12 weeks, five strains plus inulin, with a significant result on the primary endpoint (Perraudeau et al., 2020). It also ran a four-month, ten-person detection study showing its strain turns up in stool during supplementation, listed on its own clinical-trials page. No other brand here has anything comparable, and that is worth paying for if a research program is what you are buying.

Where it falls short: that trial tested the five-strain product, not this single-strain capsule, and 100 million live cells is one percent of the pasteurized dose used in the Depommier trial. You are paying $1.80 a day for live cells in a category where the one human trial found the dead cells did the work.

Best Budget Pasteurized Akkermansia: Toniiq Akkermansia 50

Akkermansia: 50 billion TFU per 2-capsule serving, heat-treated (pasteurized). No strain code printed. Also in the capsule: nothing else active. Form: two delayed-release capsules daily; "for best results, store in the refrigerator." Servings: 30 per 60-count bottle. Price: $24.97 one-time, $21.22 on subscription. Cost per day: $0.83. Testing: a downloadable COA is linked on the product page, plus a five-point per-lot program covering identity, assay, microbiology, and heavy metals; GMP certified. Verified October 14, 2026.

This is the cheapest way on the list to get a high pasteurized count, and it is the only competitor that links an actual COA from its product page. If you have read the criteria and want Akkermansia by itself, this is the sensible default.

Where it falls short: 50 billion TFU per day is above the 34 billion cells per day ceiling EFSA reviewed, so it is a dose no safety review or human trial has looked at. There is no reason to think a pasteurized ingredient turns risky at that count, but more is not the studied dose. And a pasteurized product that asks for refrigeration is a small puzzle. The cells are not alive, so what is the fridge protecting?

Best Named-Strain Pasteurized Pick: Metagenics UltraFlora Healthy Weight with Akkermansia

Akkermansia: 176 mg pasteurized Akkermansia muciniphila MucT, 30 billion TFU, per capsule. Also in the capsule: green tea leaf extract 250 mg (100 mg EGCG), chromium picolinate 200 mcg, vitamin D 5 mcg, riboflavin 0.21 mg. Form: one capsule daily; no refrigeration required. Servings: 30. Price: $69.95, and out of stock on the brand's own store on October 14, 2026. Cost per day: $2.33. Testing: the brand states it guarantees the labeled TFU count through expiration; no per-lot document linked. Verified October 14, 2026.

MucT is the type strain of the species, the original 2004 isolate (ATCC BAA-835), and the strain whose Amuc_1100 protein was characterized in the mechanism papers above. Metagenics is the only brand here that names it, at 30 billion TFU, three times the human-trial dose and inside the EFSA ceiling. If "which strain is this" is your first question, this is the product that answers it.

Where it falls short: $2.33 a day is the highest cost on the list, part of that pays for green tea extract and chromium you may not want, the brand sells mainly through practitioners, and on the day we checked you could not buy it from the brand's store at all.

Best Budget Pick If You Want a Live Claim: Double Wood Akkermansia

Akkermansia: "Akkermansia 50 mg (Akkermansia muciniphila AH39) 5 billion TFU" per 2-capsule serving, per the label image on the brand's page. The product copy says third-party testing at 50 billion CFU per gram shows "the entire 5 billion of TFU is alive." Also in the capsule: rice flour, hypromellose, maltodextrin. Form: two capsules daily; no storage instruction on the page. Servings: the copy says 60 capsules and 30 servings; the label image on the same page says 60 servings. We priced it at 30. Price: $24.95. Cost per day: $0.83. Testing: "third-party tested" and "GMP certified facility" badges; no COA linked. Verified October 14, 2026.

It names a strain, AH39, which most of the list does not, and it costs the same per day as Toniiq. For someone who specifically wants a live-cell claim at the lowest price, nothing else here comes close.

Where it falls short: the label counts in TFU, a total-cell unit, while the copy claims viability with a CFU-per-gram figure that is not on the label; the serving count disagrees between the label image and the description; there is no storage guidance for a product the brand says is alive; and 5 billion is half the trial dose. The claim may well be true. The documentation on the page does not let you check it.

Side-by-side

All prices are one-time purchase prices from each brand's own site on October 14, 2026.

Red flags

A count with no unit, or the wrong unit

"5 billion Akkermansia" with no CFU, AFU, or TFU after it tells you nothing. A TFU count on a product sold as live tells you the brand either does not understand its own unit or hopes you will not. Ask which it is.

No strain designation

"Akkermansia muciniphila" is a species. The trials were run on specific strains. A label that stops at the species name is asking you to assume the product matches the research. Three of the five products here do that, including ours, which is why we say so in our own entry instead of pretending the criterion does not apply to us.

A dose with no trial behind it, in either direction

One percent of the studied dose is a different product from the studied dose. So is five times it. Neither is disqualifying, but a brand that does not tell you where its number sits relative to 10 billion cells a day is leaving out the one comparison that matters.

"Clinically proven" with no study you can click

Pendulum's trial has a PubMed record. Our testing has a lab page. If a product page says "clinically studied" and the only link goes to a blog post, there is no study. What we learned testing 44 supplements bought on Amazon covers what happens when nobody checks.

Refrigeration instructions that do not match the product

A pasteurized product that needs a fridge, or a live product with no storage guidance at all, is a sign the label and the formulation were written by different people. Two products on this list have one of those problems.

Per-bottle prices that hide the per-day math

$24.95 for 60 capsules sounds like 60 days until you read "serving size: 2." Every pick above is priced per day for that reason. Do the division before you compare. And buy from the brand's own site: Pendulum's wholesale terms, posted on its site, prohibit resale on Amazon and Walmart, so a Pendulum listing there did not come through an authorized channel.

Who should skip this entirely

If you are pregnant or breastfeeding, skip pasteurized Akkermansia. The one regulator that has reviewed it excluded you from the authorization, not because of a finding but because nobody has tested it in you. If you are severely immunocompromised, talk to your clinician before any live bacterial product. If your actual complaint is bloating after meals or trouble with specific foods, a conventional multi-strain probiotic or a digestive enzyme has more direct human data than Akkermansia does and costs less. And if you eat a lot of fiber and not much ultra-processed food, you are already doing the thing the animal work suggests raises Akkermansia on its own: in the 2013 mouse study, prebiotic feeding restored the bacterium's abundance before anyone fed the mice the bacterium itself. What the gut microbiome does and how diet moves it is the better starting point for most healthy adults.

Next reads

• How Akkermansia makes a probiotic better: the mechanism case for pairing a postbiotic with live strains, in plain language.
• Seed DS-01 versus our 3-in-1 Probiotic Fusion: a head-to-head with the best-known synbiotic that does not contain Akkermansia.
• How we rank our own products against competitors: the Urolithin A guide, where we gave ourselves one pick out of five.

*These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.