Levels is no longer accurately described as a $100-plus monthly prescription-CGM bundle. In 2026, the company sells a lower-cost app membership and treats sensors as optional add-ons. Its most accessible sensor option is Dexcom Stelo, an over-the-counter CGM for eligible adults that does not require a prescription.

That makes Levels easier to try, but it does not settle the bigger question: does a healthy person gain enough useful information to justify wearing and paying for a CGM? The answer depends on the goal. A short, structured experiment may reveal repeatable glucose responses and prompt useful behavior changes. Open-ended monitoring can also create cost, noise, anxiety, and false confidence.

This July 2026 refresh corrects outdated pricing, sensor, prescription, accuracy, and weight-loss claims. It does not treat CGM data as a diagnosis or a substitute for medical care.

What Levels sells now

Levels is primarily an analysis and coaching app. It can combine food logs, activity, sleep, uploaded laboratory results, and compatible CGM data into meal scores, trends, and behavior prompts.

Levels currently lists:

• membership: $15 monthly or $80 annually;
• Stelo subscription: $89 per two-sensor shipment;
• one-time Stelo pack: $99 for two sensors;
• optional lab panels and nutrition sessions: priced separately; and
• Dexcom G7: available through a separate prescription-facilitation route.

Each Stelo sensor is intended to last up to about 15 days, so two sensors represent roughly a month of continuous wear. Levels notes that not every sensor lasts the full stated period. Hardware, shipping, tax, app renewal, and any optional services should all be included when comparing the real cost.

Check current Levels membership and sensor terms before purchasing.

The important access correction: OTC CGM exists

The original version of this review said CGMs in the US require a prescription and that consumer services solve this through telehealth. That is outdated.

The FDA cleared Dexcom Stelo as an over-the-counter integrated CGM in 2024. It is intended for adults 18 and older who do not use insulin, including people without diabetes who want to understand how diet and exercise relate to glucose. Abbott’s Lingo and Libre Rio also received OTC clearances for specified non-insulin-using adult populations.

OTC does not mean “appropriate for everyone” or “diagnostic.” Read the exact device label. People using insulin, people with problematic low glucose, pregnant users, or anyone making treatment decisions need clinical guidance and a device appropriate to their situation.

What a CGM actually measures

A CGM sensor estimates glucose in interstitial fluid, not directly in a blood sample. Readings can lag behind blood glucose, especially when glucose is changing quickly. Compression during sleep, a newly inserted sensor, exercise, hydration, medication interference, and sensor problems can also affect the trace.

The useful information is usually a pattern:

• how glucose changes after a repeatable meal;
• whether a walk after eating changes the curve;
• how sleep loss or a hard workout coincides with a different response;
• how long a rise persists; and
• whether the pattern repeats under similar conditions.

One isolated “spike” is not a diagnosis, moral judgment about a food, or proof that the meal caused long-term harm. The app’s score is a product-generated interpretation, not a universally accepted clinical endpoint.

What evidence supports—and what it does not

CGM has strong clinical value for many people with diabetes, especially those using insulin or therapies that can cause hypoglycemia. The evidence is much less mature for broad optimization in otherwise healthy people.

Personal glucose responses to identical foods can vary, and feedback may help some users connect meals and activity with repeatable patterns. But a device can also encourage overreaction to normal physiology. Adding protein, fat, or fiber may flatten a graph while increasing total calories; avoiding nutritious fruit or legumes merely to chase a score is not automatically healthier.

The American Diabetes Association’s 2026 diagnostic guidance states that evidence is insufficient to use CGM for screening or diagnosing prediabetes or diabetes. If symptoms, family history, or a CGM pattern raises concern, the next step is appropriate laboratory testing—not self-diagnosis from an app.

Claims that CGM itself causes weight loss are also too strong. Feedback may support behavior change, but the sensor is an information tool. Sustainable food intake, activity, sleep, medication when indicated, and clinical care determine outcomes.

Who may get useful value

A time-limited Levels trial is most reasonable for an adult who:

• has one or two specific behavior questions;
• is willing to repeat comparable meals or activities;
• understands that the trace is not a diagnosis;
• can act on simple findings without becoming restrictive; and
• can afford the experiment without treating it as essential care.

Examples include comparing a normal lunch with the same lunch followed by a walk, testing a breakfast change across several similar days, or observing fueling during a consistent training block.

Levels is less compelling for someone who:

• already follows a stable, balanced routine and has no clear question;
• feels anxious or compulsive around food and health data;
• wants the app to diagnose unexplained symptoms;
• would change medication based on an OTC trace; or
• expects a permanent “perfect” glucose line.

A better two-to-four-week experiment

A Stelo sensor is designed for up to about 15 days. That makes one sensor a natural two-week experiment; the current $89 subscription shipment or $99 one-time pack includes two sensors and can extend the same protocol to roughly four weeks. Write down the question before applying the first sensor and keep the experiment small enough to interpret.

1. Use days 1–3 as a baseline while eating and moving normally.
2. Pick no more than two changes, such as a 20-minute post-meal walk or a repeatable breakfast with a measured fiber addition.
3. Repeat each condition on at least two comparable days when practical.
4. Track hunger, energy, sleep, gastrointestinal comfort, and training—not just the glucose curve.
5. Avoid labeling single foods “good” or “bad” from one response.
6. Stop if the data increases anxiety or drives unnecessary restriction.
7. Bring persistent or concerning patterns to a clinician for proper testing.

For many people, one Stelo cycle provides enough information to decide whether Levels changes behavior. Do not buy the second sensor or renew the app merely to collect more graphs; continue only when the first cycle produced a specific question worth retesting. Continuous year-round wear is not required to benefit from a focused experiment.

Safety and interpretation

Follow the sensor’s current instructions for insertion, compatible phones, skin preparation, water exposure, and removal. Stop and seek advice for significant skin reaction, bleeding, signs of infection, or repeated sensor failure.

Do not use an OTC wellness CGM to make insulin, medication, or emergency-treatment decisions outside its labeling. When a reading conflicts with symptoms or seems implausible, follow the device instructions and obtain appropriate confirmatory testing. Urgent symptoms deserve medical care regardless of what the app displays.

CGM data should not reassure someone away from screening. A1C, fasting plasma glucose, or an oral glucose-tolerance test may be appropriate depending on the clinical question.

Levels versus using Stelo alone

Stelo includes its own app. Levels adds food logging, cross-signal analysis, scoring, programs, and a longer-term dashboard. The extra membership is worthwhile only if those features change decisions or make the data easier to use.

Option	Best for	Main limitation
Stelo with its native app	A lower-cost, focused glucose experiment	Less integrated food and lifestyle coaching
Levels plus Stelo	People who will use logging, programs, and combined trends	Extra subscription and more interpretation layers
Clinician-guided CGM	A medical question or treatment plan	Requires appropriate clinical access and device choice
No CGM	Most healthy people working on fundamentals	No real-time personal glucose trace

Verdict

Levels is easier to recommend as a short, question-driven experiment than as a permanent subscription for every health-conscious adult. The current lower membership price and OTC Stelo option remove much of the old access friction. The app’s logging and pattern tools may help someone run a more organized experiment than a sensor alone.

It is not a diagnostic service, a guaranteed weight-loss tool, or a prerequisite for metabolic health. If the data will not change a decision—or if it makes eating more anxious—the better choice is to skip it and focus on established basics.

Frequently asked questions

Do I need a prescription?

Not for the current Stelo OTC option if you meet its labeled eligibility. Dexcom G7 remains available through a prescription pathway.

How much does Levels cost?

As of July 23, 2026, Levels lists membership at $15 monthly or $80 annually. Optional Stelo shipments are $89 by subscription or $99 as a one-time two-pack. Verify live pricing before checkout.

Can Levels diagnose prediabetes?

No. CGM is not established as a screening or diagnostic test for prediabetes or diabetes. Use appropriate laboratory testing and clinical interpretation.

How accurate is Stelo?

It is an FDA-cleared OTC integrated CGM for its labeled population, but no sensor is perfect. Interpret repeated trends, follow the device instructions when readings and symptoms disagree, and do not extend it beyond its intended use.

Can it help with weight loss?

It may help some people change behavior, but the sensor itself does not cause weight loss and a flatter graph is not automatically a healthier diet.

Sources and verification

• Levels pricing and plans — membership and optional add-ons, updated June 18, 2026.
• Levels Stelo sensor pricing — shipment options and expected wear period, updated June 18, 2026.
• FDA Stelo clearance announcement — intended population and OTC status.
• FDA sensor-based device clearances — Stelo, Lingo, and Libre Rio regulatory records.
• American Diabetes Association, 2026 diagnosis and classification guidance — CGM is not established for screening or diagnosis.
• Zeevi et al., personalized glycemic responses, Cell (2015).

Related articles

• Oura Ring vs WHOOP
• 16:8 Intermittent Fasting Guide
• Biohacking Beginner’s Guide