Can You Buy a FreeStyle Libre Without a Prescription?

In the United States, the FreeStyle Libre 2 and Libre 3 (and 3 Plus) are prescription devices. You cannot walk into a pharmacy and buy one over the counter. Since the FDA cleared the first over-the-counter integrated continuous glucose monitors in 2024, three sensors have become available without a prescription: Abbott’s Libre Rio and Lingo, and Dexcom’s Stelo. All three use the same sensor hardware families as their prescription siblings. What differs is the firmware, the display range, and the absence of hypoglycemia alarms. If you specifically want a Libre 2 or 3, the practical route is a telehealth prescription. Several services issue one in a few minutes for a fee, and you then fill it at a pharmacy with a discount card.

The rest of this article walks through the over-the-counter options. It covers what their narrower clearances cost you and how to get usable data out of each one. It also covers what the resulting glucose trace can and cannot tell you.

What is available over the counter in the US

Three products are on the market as of this writing. They differ more in software than in hardware, so the choice largely comes down to what the app will show you and let you export.

• Dexcom Stelo. Fifteen-day wear, built on the Dexcom G7 sensor platform. It is phone-only, with no separate receiver. It is sold direct and through some pharmacies and Amazon, at roughly $99 for a two-sensor box and less on subscription.
• Abbott Lingo. Fourteen-day wear, built on the Libre 3 sensor platform and sold direct. It is priced per two-week sensor with volume discounts on 6-packs. The Lingo app converts glucose excursions into a proprietary score rather than showing you a plain mg/dL trace as the primary view.
• Abbott Libre Rio. Fourteen-day wear, aimed at adults with type 2 diabetes who are not using insulin. Its rollout has been slower and less broadly stocked than Stelo’s.

Prices move, so check the manufacturer’s direct store and GoodRx for your ZIP before relying on any number here. The useful unit for comparison is cost per day of coverage. A $99 two-pack of 15-day Stelo sensors works out to about $3.30 per day, and a $49 single 14-day Lingo sensor is about $3.50 per day. A prescription Libre 3 filled for cash with a discount card is commonly quoted in the $60–$80 range per 14-day sensor, which comes to $4.30–$5.70 per day. Costco’s member pricing is usually at or near the low end of the cash range. It is still a prescription fill rather than an over-the-counter purchase.

What you give up with the OTC versions

The clearances on the over-the-counter devices are narrower by design, and the narrowing shows up exactly where safety is at stake. These devices are not indicated for people on insulin, and they do not issue urgent low alarms. Stelo’s display range is clipped at the bottom, so a genuine hypoglycemic excursion reads as “low” rather than as a number. Lingo’s app keeps much of the raw trace behind its scoring layer. If you are doing anything where the bottom of the range matters, that clipping is a real loss of information. The absence of alarms is the reason these devices are not cleared for insulin users.

The sensor chemistry is the same wired-enzyme glucose oxidase platform across the over-the-counter and prescription lines, so the physics of measurement is identical. Both read glucose in interstitial fluid rather than blood. Both lag plasma glucose by 5 to 15 minutes, and both drift during warm-up. What differs is the algorithm, the indication, and the software.

There is one case where the choice should not be made in a store. If you use insulin or a sulfonylurea, or you are investigating symptoms that could be hypoglycemia, that decision belongs with a clinician. The over-the-counter devices are explicitly not cleared for that use. A sensor that cannot display a number below its floor is the wrong instrument for the question.

Getting the data out

Data export is where these products diverge most. This matters if you plan to analyze the trace yourself rather than only reading the app, so it is worth understanding each path before you buy.

For the prescription Libre 2 and Libre 3, the official path is LibreView, Abbott’s web console. You upload from the LibreLink app and then export from the console. That gives you a CSV file, a plain-text table with one row per reading. The columns include Device and Serial Number alongside Device Timestamp and Record Type. There is also Historic Glucose mg/dL and Scan Glucose mg/dL. Fields for notes, insulin, and carbohydrates follow. Record Type 0 is the 15-minute historic value, 1 is a scan value, and 5 and 6 are notes. The first two lines are metadata, so read the file with skiprows=1. Parse timestamps with format="%m-%d-%Y %I:%M %p", which is locale-dependent and worth checking against your own export. Libre 3 logs at 1-minute resolution to the app. LibreView still exports the 15-minute historic series in most configurations, and that is the single most common surprise for people expecting minute data.

If you want minute-level data from a Libre sensor, the community tools are Juggluco and xDrip+. Juggluco runs on Android and reads the Libre 2 and 3 Bluetooth Low Energy (BLE) broadcast directly. It also serves a local Nightscout-compatible endpoint at http://localhost:17580/api/v1/entries.json. Both tools require you to start the sensor with their app rather than Abbott’s, and both sit outside the manufacturer’s terms of use. Libre 3’s BLE payload is encrypted, so these tools depend on reverse-engineered key derivation. That derivation breaks whenever Abbott ships firmware changes, so budget for a sensor generation occasionally going dark for a few weeks.

Stelo is the most accommodating of the three. It offers in-app CSV export, and Dexcom’s developer API covers Stelo data through the same OAuth 2.0 authorization flow used for the G6 and G7. The API returns 5-minute estimated glucose values (EGVs). Each reading carries the fields systemTime and displayTime along with value and trend. If you want a programmatic pipeline without patching anything, this is the cleanest option available today. It is the main reason we default to Stelo for people who want to wear a CGM purely to generate an analyzable time series.

Lingo sits at the other end. CSV export exists, but the app is the least friendly to people who want the raw series rather than the score.

Whatever the source, normalize to a single schema early. That schema needs a UTC timestamp, a mg/dL integer, plus source device and sensor serial. The sensor serial matters because sensor-to-sensor bias is the largest single source of variance in a multi-month record. Model it as a per-sensor offset, so that you do not read a change of sensor as a change of physiology.

Failure modes you will hit

A few artifacts appear reliably enough in long CGM records that it is worth planning your analysis around them from the start. Each one has a characteristic signature and a straightforward way to handle it.

• Warm-up drift. The first 12 to 24 hours after insertion read differently from the rest of the wear. We drop the first 12 hours by default.
• Compression lows. Sleeping on the sensor produces a smooth 20–40 mg/dL dip that recovers when you roll off. It looks like nocturnal hypoglycemia without being it. Flag any sustained low between 00:00 and 06:00 with a slow, symmetric shape for manual review.
• Sensor-to-sensor offset. Two sensors worn simultaneously on the same person routinely differ by 10 to 15 mg/dL in mean. Compare within a sensor rather than across sensors.
• Population-specific accuracy loss. Sensor accuracy is not uniform across conditions. In people undergoing hemodialysis, both Dexcom and Libre sensors showed degraded accuracy, with performance varying around dialysis sessions.¹ Accuracy studies done in stable outpatients do not transfer automatically to your situation.

What the trace is good for, and what it is not

Interpreting your own data calls for some care about where the evidence comes from. Nearly all the outcome evidence for CGM comes from diabetes populations. Flash glucose monitoring produced reductions in HbA1c, the standard three-month average marker of glucose control, in type 2 diabetes managed with basal insulin. That result appeared both in real-world chart review and in meta-analysis,² and the size of the reduction in a large national cohort depended strongly on baseline HbA1c and on how often people scanned.³ CGM-derived metrics such as time in range and glycemic variability associate with clinical parameters in type 2 diabetes.⁴ Whether those metrics predict macrovascular and microvascular complications is still being studied in dedicated retrospective cohorts.⁵

None of that establishes what a time-in-range number means for a metabolically healthy person with no diagnosis. The productive way to use your own record is as a personal, within-subject measurement instrument. It is good at telling you how your own glucose responds to a specific meal, a specific sleep pattern, or a specific exercise timing. That holds when you test it repeatedly on yourself. It is bad at telling you where you stand against a population.

One exception deserves naming. If your fasting values run persistently high, or you see repeated true lows that are not compression artifacts, that is data for a clinician and a venous test rather than material for a self-directed experiment. CGM has been used intraoperatively to track glucose during insulinoma surgery,⁶ a reminder that these sensors are deployed in serious clinical settings under supervision and do not substitute for it.

Questions people also ask

Do I need a prescription to buy a Libre sensor? For FreeStyle Libre 2, 3, and 3 Plus in the US, yes. For Libre Rio and Lingo, no. Outside the US the rules differ by country, and Abbott sells some Libre products direct to consumers.

What is the cheapest OTC CGM? On a per-day basis, Stelo on subscription and Lingo’s multi-pack pricing land close together, around $3 to $3.50 per day. Stelo’s 15-day wear against Lingo’s 14-day is a small edge. Stelo’s API access is a larger one if you want the data programmatically.

What is cheaper, Libre or Dexcom? Prescription Libre is usually cheaper per sensor than prescription Dexcom G7 at cash price. Among the over-the-counter devices the gap narrows to near nothing.

How much is a Libre sensor out of pocket? Commonly $60 to $100 per 14-day sensor at cash price with a discount card, varying by pharmacy and ZIP. Abbott has run first-sensor-free programs for new prescription users.

How do I get Libre 2 sensors for free? Manufacturer starter programs cover a first sensor for eligible new users with a prescription, and some insurers cover sensors under the pharmacy benefit. There is no general free supply.

Woolf Software builds longitudinal molecular profiles of individuals: whole-genome sequencing, RNA sequencing, proteomics, blood biomarkers, and continuous glucose data, integrated into one model of you. Build your profile.

Footnotes

1. Parizad Avari, Wenxi Tang, Narvada Jugnee, et al. The Accuracy of Continuous Glucose Sensors in People with Diabetes Undergoing Hemodialysis (ALPHA Study). Diabetes Technology & Therapeutics, 2023. https://doi.org/10.1089/dia.2023.0013 ↩

2. Anders L Carlson, Timothy Dilon Daniel, Andrea DeSantis, et al. Flash glucose monitoring in type 2 diabetes managed with basal insulin in the USA: a retrospective real-world chart review study and meta-analysis. BMJ Open Diabetes Research & Care, 2022. https://doi.org/10.1136/bmjdrc-2021-002590 ↩

3. A. Lameijer, M.J. Fokkert, M.A. Edens, et al. Determinants of HbA1c reduction with FreeStyle Libre flash glucose monitoring (FLARE-NL 5). Journal of Clinical & Translational Endocrinology, 2020. https://doi.org/10.1016/j.jcte.2020.100237 ↩

4. Purvi Chawla, Alpana Sowani, Rakesh Parikh, et al. Association between continuous glucose monitoring derived metrics and clinical parameters in Indian people living with type 2 diabetes. Primary Care Diabetes, 2025. https://doi.org/10.1016/j.pcd.2025.06.004 ↩

5. Ramzi A Ajjan, Tadej Battelino, J Seufert, et al. Do continuous glucose monitoring (CGM) metrics predict macrovascular and microvascular complications in diabetes? The FACULTY protocol of a retrospective real-world cohort study. BMJ Open, 2025. https://doi.org/10.1136/bmjopen-2024-085961 ↩

6. M. Magliozzo, A. Tumminia, M. L. Arpi, et al. Intraoperative intermittently scanned continuous glucose monitoring in the management of patients with pancreatic insulinoma. Journal of Endocrinological Investigation, 2024. https://doi.org/10.1007/s40618-024-02472-6 ↩