Levels vs Lingo vs Stelo: Continuous Glucose Monitor Comparison for Biomarker Tracking 2026

Educational content for biomarker interpretation. Not medical advice.

What Changed: Non-Prescription CGM Goes Mainstream

For most of the past decade, continuous glucose monitors (CGMs) were prescription-only medical devices intended primarily for type 1 and insulin-treated type 2 diabetes management. That changed substantially in 2024 and 2025 with three consumer-direct products gaining regulatory clearance for over-the-counter sale in the United States:

• Stelo by Dexcom (FDA-cleared March 2024), sensors based on the G7 platform, sold directly to consumers without a prescription
• Lingo by Abbott (consumer launch 2024), sensors derived from the FreeStyle Libre 3 platform, marketed to non-diabetic users
• Levels Health, a behavioural and analytics platform that uses Dexcom and FreeStyle Libre sensors and provides a paid software experience, available with a telehealth prescription in the US

This brief compares the three for the use case of biomarker tracking, periodic or continuous monitoring as part of a broader functional medicine or quantified-self protocol, rather than diabetes management.

For the foundational interpretation framework that applies to any glucose data, see our pieces on how to interpret blood test results, HbA1c optimal ranges and functional targets, and fasting insulin: the missing test on why glucose-only data is incomplete without insulin context.

Quick Comparison Table

Pricing and availability are accurate to 2026-05-23 and will shift as these products expand to new markets.

Levels Health

What it is: Levels is fundamentally a software experience layered over consumer or prescription CGM hardware. Users pay a subscription that includes sensors (Dexcom G7 or, historically, FreeStyle Libre) shipped to them, plus an app that scores meals, identifies glucose patterns, and provides coaching content.

Strengths:

• The behavioural-change focus is the most developed of the three. The app explicitly grades meal responses on a 1 to 10 scale, integrates exercise and sleep data, and prompts experiments.
• Strong data-export options including CSV and a developer API for users who want to integrate glucose data into other tools.
• Includes a periodic 1:1 coaching component.

Limitations:

• Most expensive of the three by a meaningful margin.
• Requires a telehealth prescription in the US, which adds friction.
• Australian access remains limited, Levels does not officially ship to Australia, though some users have reported workarounds.

Best for: users who want a structured, coaching-driven introduction to metabolic data and are willing to pay a premium for the curated software experience.

Lingo (Abbott)

What it is: Lingo is Abbott's consumer-direct CGM line, sold without a prescription in markets where it has launched. Sensors are derived from the FreeStyle Libre 3 platform, the same underlying technology used in Abbott's prescription CGMs, with a consumer-targeted app overlay.

Strengths:

• No prescription required where launched. Friction-free purchase path.
• Mature underlying sensor technology, FreeStyle Libre platforms have years of real-world validation.
• Engaging app experience built around metabolic "scores" and daily streaks.

Limitations:

• App is more locked-down than competitors. Data export is limited to the app itself in some configurations.
• Coaching and behavioural content is less developed than Levels.
• Australian launch timing is not yet confirmed at the time of writing.

Best for: users who want straightforward consumer-direct CGM access with a simple, engagement-oriented app, and who do not need extensive data export capabilities.

Stelo (Dexcom)

What it is: Stelo is Dexcom's consumer-direct line, FDA-cleared for over-the-counter sale in March 2024. Sensors share the G7 platform underlying Dexcom's prescription CGMs but are repositioned for non-diabetic metabolic awareness.

Strengths:

• Lowest entry cost of the three.
• 15-day sensor life, longest in the comparison.
• Dexcom platform compatibility means data can integrate with the broader Dexcom ecosystem for users who later transition to prescription Dexcom use.

Limitations:

• App experience is the most basic of the three. Pattern detection is present but coaching content is minimal.
• Not yet available in Australia or most non-US markets.
• Limited data export options compared to Levels.

Best for: users who want raw glucose data with the lowest financial commitment and minimal app friction, and who plan to interpret the data themselves rather than relying on app-driven coaching.

Sensor Accuracy: What the Data Shows

All three products use FDA-cleared sensor technologies. Published Mean Absolute Relative Difference (MARD) values, the standard CGM accuracy metric, fall in the range of 8 to 10% for Dexcom G7 and FreeStyle Libre 3 in adult populations. This is broadly equivalent across the three products at the sensor level.

Practically, CGM accuracy is most variable:

• During rapid glucose excursions (post-meal, post-exercise)
• In the first 12 to 24 hours after sensor insertion (warm-up period)
• At low glucose values (under 70 mg/dL or 3.9 mmol/L), where sensor lag is more pronounced
• In the presence of certain medications and supplements (high-dose vitamin C, paracetamol/acetaminophen at higher doses for some sensor types)

For periodic biomarker tracking, these accuracy considerations matter less than they do for insulin-dose decisions in type 1 diabetes. A trend pattern is more interpretively important than any single reading.

Use Case 1: Casual Metabolic Awareness

For a user who wants to understand how meals, sleep, and exercise affect glucose over a short observation period (typically 2 to 4 weeks), any of the three products will provide adequate information. The choice reduces to:

• Cost-conscious: Stelo
• Engagement-driven app: Lingo
• Coaching and structured experiments: Levels

A short observation window is often sufficient for the primary insight, most users discover their major glucose drivers within the first two weeks of wearing a CGM.

Use Case 2: Peptide Research Monitoring

For users self-tracking glucose responses during peptide-related research protocols (particularly involving GLP-1 class compounds like semaglutide or tirzepatide, which are well known to influence postprandial glucose patterns) CGM data adds an objective measurement layer that paired blood glucose and HbA1c testing alone cannot provide.

Practical considerations specific to this use case:

• A CGM captures the temporal shape of glucose responses to meals, which is exactly what GLP-1 agonists are expected to flatten via slowed gastric emptying and postprandial insulin sensitisation
• HOMA-IR calculated from periodic fasting glucose and insulin (see our HOMA-IR insulin resistance calculator) provides complementary insulin-sensitivity context that CGM data alone does not
• Baseline glucose response patterns captured before initiating any research peptide protocol provide the most useful comparison data

The data-quality value of paired CGM monitoring during any research peptide protocol is highest when baseline glucose patterns have been captured for at least two weeks before the protocol begins, and when meal and activity logs are kept alongside the glucose stream so post-hoc analysis can separate compound effects from confounders.

For users who want a more comprehensive baseline biomarker workup before initiating any monitoring protocol, our private blood test Australia guide covers the available providers and panel options.

Use Case 3: Athletic and Performance Optimisation

CGM data has growing adoption among endurance athletes and strength-trained populations for fuelling strategy refinement. The use case is less about disease screening and more about operational glucose data for training and competition.

For this audience:

• Levels offers the most developed exercise and performance integration
• Stelo's lowest cost makes it viable for the long observation periods (months) that athletic optimisation typically requires
• Sensor placement matters more for athletic users, sensors on the upper arm typically survive contact and movement better than abdomen placements

Caveats: What CGMs Are Not Good For

Three honest limitations to set against the marketing claims:

1. CGMs do not measure insulin. Glucose data alone is incomplete. Two users with identical glucose curves can have very different insulin profiles, and the insulin profile is often the more informative one for metabolic health. Fasting insulin, HOMA-IR, and C-peptide testing remain essential complements to CGM data.

2. CGM data quality degrades at glucose extremes. Very low or very rapidly changing glucose values are where sensor lag and noise are highest. For users with normal glucose ranges, this matters less than it does for type 1 diabetes management, but it is worth understanding.

3. CGMs do not replace a structured biomarker panel. A comprehensive metabolic biomarker workup includes apolipoprotein B (see our ApoB optimal ranges guide), liver enzymes, kidney panel, inflammation markers, and a comprehensive lipid panel, none of which CGM data substitutes for.

Australian Access Reality

As of 2026-05-23, all three products have limited or no official Australian availability:

• Stelo: US only as of writing. No published Australian launch timeline from Dexcom.
• Lingo: Launched in several markets but Australian timing remains unconfirmed by Abbott Australia.
• Levels: Does not officially ship to Australia. Some Australian users have reported workarounds, but these are outside Levels's official service.

Australian users seeking CGM access today still primarily go through prescription pathways for FreeStyle Libre or Dexcom G7 sensors, typically via GP or endocrinologist referral. Some private health funds provide partial subsidies for non-diabetic CGM use; PBS subsidy for non-diabetic CGM is not currently available.

Summary

For 2026, the consumer CGM landscape offers three meaningfully different products. Levels competes on behavioural coaching and data sophistication at a premium price. Lingo offers consumer-direct access with engaging gamification at a moderate price. Stelo offers the lowest entry cost with the longest sensor life and the most basic app experience.

For Australian users, the practical reality is that none of the three is yet officially available domestically. The decision for now is whether to pursue prescription CGM access through standard medical channels, or to wait for consumer-direct launches that may arrive in 2026 or 2027.

For biomarker tracking purposes, CGM data is most useful as one layer in a broader monitoring approach. Pair glucose curves with fasting insulin, HOMA-IR, HbA1c, and lipid biomarkers for a complete metabolic picture rather than treating CGM data as a stand-alone diagnostic.