Independent since 2023  ·  No affiliate links.  ·  We buy every subscription ourselves. New York, NY Issue No. 148
Home / Head-to-head / Levels vs Nutrisense
Head-to-headMetabolicCGM

Levels vs Nutrisense: two ways to spend $1,000 on your own blood sugar

Both wrap the same continuous glucose sensors in software. One sells you a system to learn; the other sells you a dietitian to interpret. The gap between their sticker prices is about $3,300 over two years — and there's a third option neither of them puts on the pricing page.

By Marcus Oyelaran Published 12 Aug 2026 Updated 1 Sep 2026 Next re-test Jan 2027 16 min read
Short answer

Levels is the better software and far cheaper; Nutrisense is worth it only if you'll use the dietitian. Levels runs $80–$1,999 a year depending on tier; Nutrisense starts at $179 a month ($2,148 a year), with coaching often billed to insurance. If you're metabolically healthy and self-directed, the cheapest honest path is an over-the-counter sensor plus Levels' $80/year app tier.

71
Applandica score

Levels

$80–$1,999/yr across three tiers · sensors bundled or BYO

Better software, better value
66
Applandica score

Nutrisense

From $179/mo · sensors included · dietitian often $0 via insurance

Better human support

The 60-second scorecard

Green cells are the winner on that row. Pricing verified against both companies' own sites on 1 September 2026.

What you care aboutLevelsNutrisense
Cheapest way in$80/yr (app only, BYO sensor)$179/mo
Mid tier with sensors$499/yr$2,148/yr
Works without a sensorYesNot meaningfully
Bring your own sensorSupportedLimited
Registered dietitianNoYes — often $0 out of pocket via insurance
Curve annotation & food scoringBest in categoryFunctional
Explains why a curve movedPartlyPartly, plus a human who can
Raw data exportComplete CSVPartial
Onboarding for non-diabeticsAssumes some contextExplains the basics well
Insurance / HSA-FSA pathwayHSA/FSA eligibleCoaching billed to insurance
24-month cost, mid tier$998$4,296

What does each one actually cost in 2026?

Both companies restructured pricing in the past year, and most comparison articles you'll find are quoting numbers that no longer exist. Here is what their own sites showed on 1 September 2026.

PlanMonthlyAnnualSensors included24-month total
Levels — Build Your System$15$80None (BYO or add on)$160 + sensors
Levels — Core$41$4991 month of CGM$998
Levels — Complete$167$1,9992 months of CGM$3,998
Nutrisense — CGM programmefrom $179$2,148Yes, ongoing$4,296

List prices, no promo codes applied. Nutrisense was advertising 15% off at the time of checking, which brings the monthly to about $152. Dietitian and stress-management sessions are billed as $0 out of pocket for most members with eligible insurance.

$3,298

the two-year difference between Levels Core and Nutrisense at list price. That is not a feature gap — it is a staffing model. You are paying for a person, and whether that's good value depends entirely on whether you talk to them.Applandica pricing database · verified 1 Sep 2026

The third option neither of them advertises

Over-the-counter glucose sensors no longer require a prescription in the US. Dexcom Stelo runs roughly $89–99 a month; Abbott Lingo is about $49 per 14-day sensor. Levels' cheapest tier explicitly supports bringing your own sensor and works without one at all.

So the cheapest legitimate configuration for a metabolically healthy adult who wants to run a two-month experiment is:

  • Levels "Build Your System" at $80 for the year, plus
  • two months of an OTC sensor at roughly $100–$180 total,
  • total: about $180–$260 — against $358 for two months of Nutrisense, or $998 for two years of Levels Core.

Nobody markets this configuration because it is the least profitable one. It is also, for most readers of this site, the correct one. You are not buying a lifelong subscription to glucose; you are buying a short, sharp experiment. Everything after month three is repetition.

Does CGM actually do anything for people without diabetes?

Honestly? The evidence is thinner than either company's marketing implies, and the useful findings point somewhere cheap.

A 2025 systematic review in Cureus pooled seven studies covering roughly 1,127 non-diabetic participants using CGM to guide lifestyle change. Its conclusion was carefully hedged: CGM "shows promise for personalising lifestyle interventions", but direct evidence on traditional cardiovascular risk factors was limited, and there is nothing yet on hard cardiovascular endpoints.

The specific findings inside that review are more interesting than the headline:

  • Timing beats composition. A 20-minute walk taken before your individual post-meal glucose peak cut the four-hour incremental glucose AUC by 0.6 mmol·L⁻¹·h, and insulin and C-peptide by 28.7% each.
  • Activity breaks work. Regular movement breaks lowered fasting glucose by 0.34 ± 0.37 mmol/L and cut daily glucose variability by about 2%.
  • Physiology shifts with life stage. Post-menopausal participants showed a 42% higher two-hour glucose AUC and greater variability — a reminder that "your spike" is not a moral result.

Read that list again. The best-evidenced action to come out of consumer CGM research is go for a walk after you eat — and you do not need a $4,296 subscription to do it. What the sensor buys you is knowing when your personal peak lands, which is genuinely useful information, and which you can learn in about six weeks.

The real difference: a course or a coach

After eight weeks with both, the distinction that mattered wasn't software quality. It was whether you want to learn a skill or delegate it.

Levels is built on the assumption that you'll eventually read your own curves and stop paying. The annotation is better, the export is complete, and the $80 tier exists precisely because the company expects you to graduate. Nutrisense is built on the assumption that you won't, and staffs accordingly — a registered dietitian, weekly check-ins, and an insurance pathway that makes the human part cheap even when the programme isn't.

Levels is a course. Nutrisense is a coach. Both are expensive ways to discover that oats spike you and your friend's don't.

One structural complaint that applies to both: neither will show you sensor error. Consumer CGMs typically report a mean absolute relative difference around 8–10%, which means a "spike" from 95 to 140 mg/dL carries an error bar wide enough to matter. Both apps draw the curve as though it were traced from a lab analyser. It isn't.

Marcus's take

Senior Reviewer — Metabolic · RD · eight weeks on each

I came into this category sceptical and left more sceptical, which is not the arc I expected. Not because the sensors are bad — they're remarkable bits of engineering — but because of what the software does to a normal person's post-meal rise.

Six of the eight metabolic apps I onboarded this year rendered an entirely normal physiological response in alarm red. If you are metabolically healthy, going from 90 to 140 mg/dL after a bowl of rice is your pancreas working, not a wound. An app that colours that red isn't giving you information; it's giving you a mood, and charging a subscription for it.

So my honest recommendation for most people asking "Levels or Nutrisense?" is: neither, at least not at those tiers. Buy two months of an over-the-counter sensor, pair it with the cheapest Levels tier, learn three things — where your personal peak lands, which two meals actually move you, and what a post-meal walk does — then cancel and keep the walk. That's about $200 and it captures most of the available value.

Where I'd genuinely spend the money on Nutrisense: if you have real dysregulation, a diagnosis in the family, or you've tried self-directed change three times and it hasn't held. Accountability is the bottleneck for more people than data is, and a dietitian your insurance partly covers is a much better buy than another dashboard. Just go in knowing you're hiring a person, not a technology.

Who should buy which

Buy Levels if…

  • You're analytical and self-directed
  • You want complete data export
  • You plan to cancel after 2–3 months
  • You already have or can buy an OTC sensor
  • $80/year is the budget

Buy Nutrisense if…

  • Accountability, not data, is your bottleneck
  • Your insurance covers the coaching
  • You want a registered dietitian reading your curves
  • You're new to glucose and want the basics explained

Buy neither if…

  • You're metabolically healthy and just curious — buy a sensor for six weeks instead
  • You have a history of disordered eating; food scoring is a real risk here
  • You want a diagnosis — see a clinician, not an app

A safety note we won't bury. Food-scoring apps and continuous self-surveillance are a poor fit for anyone with a history of disordered eating, and both products are capable of turning an ordinary meal into a failed test. If that's you, this whole category is worth skipping — and no score we publish should override that.

Frequently asked questions

Is Levels or Nutrisense cheaper?

Levels, substantially. Levels ranges from $80/year (app only, bring your own sensor) to $1,999/year for the top tier. Nutrisense starts at $179/month, about $2,148/year. Over 24 months at mid tier that's $998 for Levels versus $4,296 for Nutrisense.

Do I need a prescription for a CGM in 2026?

Not in the US for over-the-counter sensors. Dexcom Stelo (about $89–99/month) and Abbott Lingo (about $49 per 14-day sensor) are sold without a prescription and are aimed at people without diabetes.

Does wearing a CGM improve health if I'm not diabetic?

The evidence is early. A 2025 systematic review of seven studies (~1,127 non-diabetic participants) found CGM "shows promise" for personalising lifestyle change but noted limited evidence on cardiovascular risk factors and none on hard endpoints. The best-supported actions it surfaced — walking before your glucose peak, taking activity breaks — do not require a sensor to perform, only to time precisely.

Can I use Levels with my own sensor?

Yes. Levels' entry tier is explicitly designed for bringing your own sensor, and the app functions without one. That combination is the cheapest legitimate way into the category.

Does insurance cover any of this?

Nutrisense advertises dietitian coaching and stress-management sessions at $0 out of pocket for most members with eligible insurance, which meaningfully changes its value calculation. Levels is HSA/FSA eligible but doesn't bill insurance for coaching, because there isn't any.

How accurate are consumer CGM sensors?

Typical reported mean absolute relative difference (MARD) is around 8–10% against reference measurement, and accuracy degrades at the extremes and in the first hours after insertion. Neither app displays that uncertainty, so treat individual readings as estimates and pay attention to patterns instead.

Sources

  1. Levels — plans and pricing (Build Your System $15/mo or $80/yr; Core $41/mo or $499/yr; Complete $167/mo or $1,999/yr). Verified 1 Sep 2026
  2. Nutrisense — CGM programme pricing (from $179/month; dietitian and stress-management sessions $0 out of pocket for most insured members). Verified 1 Sep 2026
  3. Use of Continuous Glucose Monitoring in Non-diabetic Individuals for Cardiovascular Prevention: A Systematic Review. Cureus, 2025 · 7 studies, ~1,127 participants. Accessed 1 Sep 2026
  4. Zhang et al. 2021, via the review above — pre-peak walking reduced 4-hour glucose AUC by 0.6 mmol·L⁻¹·h and insulin/C-peptide by 28.7%.
  5. Dexcom Stelo and Abbott Lingo over-the-counter sensor pricing. Verified 1 Sep 2026

Marcus Oyelaran

Senior Reviewer — Metabolic

Registered dietitian turned product critic. Covers CGM apps, nutrition logging and anything that scores your lunch. Deeply suspicious of food scores, including the good ones.

How we got these. Both programmes purchased at retail with our own money and run for eight weeks each, one tester per product, then swapped. No advance notice to either company, no copy approval, no affiliate relationship. Full methodology.