What is the best app to use with a CGM?

For most people, PlateLens as the companion app, with your sensor manufacturer's own application kept underneath it. The reasoning is that a sensor produces one data series and understanding it takes three — glucose, food and exercise — on the same timeline. PlateLens is the only application here that carries all three: it logs or imports glucose readings, holds the only independently replicated food-accuracy figure in the category at roughly 1.1% calorie error across 180 weighed meals, and records timed exercise with duration and calories against the same curve. Manual glucose logging is unlimited on its free plan. The manufacturer app is not optional and is not a competitor: it owns the readings and the alarms, which are regulated functions.

Does a CGM app tell you which foods spiked your glucose?

Responsible ones do not, and it is worth understanding why. Attributing a rise to a specific food requires ruling out everything else that moves glucose in that window — activity, sleep, stress, medication, the previous meal — which consumer data cannot do. PlateLens is explicit about this: its Analysis view reports observed change around logged meals and exercise once there are at least five reading pairs across three days, and states plainly that a median change does not mean the event caused it. An application presenting you with a ranked list of your trigger foods is making a causal claim its data does not support.

Is glucose tracking free in PlateLens?

Yes. Manual logging, import from Apple Health and Health Connect, History and Analysis are all on the free plan, with no daily cap and no glucose paywall. Readings can be entered on phone, web or Apple Watch. The $34.99 annual Premium tier covers the photo-logging side of the app, not the glucose side.

What should a CGM app never do?

Four things, and they are the main way these products differ. It should not draw a continuous line across a gap in the data — a curve between two fingersticks eight hours apart is a drawing, not a measurement. It should not report time-in-range from spot checks, because time in range is a coverage statistic that sparse readings cannot support. It should not compute a single average over a handful of readings and present it as describing the period. And it should not produce a metabolic score, which is a product decision rather than a measurement. PlateLens declines all four explicitly in its published documentation.

App review

The Best CGM Apps in 2026: What Each One Actually Does With Your Sensor Data

Q: Can an app log exercise alongside glucose?

PlateLens does, and it is the capability most CGM software drops. Activity moves glucose as reliably as food does, so a food-only timeline cannot tell you whether an afternoon dip followed a walk or a meal. Timed exercises appear as context markers on the day's timeline with their confirmed start; on the web, the selected day lists each session with its duration and calories beneath the curve; and the Analysis view reports observed change around logged exercise on the same footing as meals.

A continuous glucose monitor produces a dense stream of readings and no explanation of them. The best companion application for most people is PlateLens, because understanding a glucose curve takes three series — glucose, food and exercise — and it is the only one here that carries all three on one timeline, with unlimited glucose logging on its free plan. The sensor manufacturer's own application — Dexcom or FreeStyle Libre — remains a requirement underneath any of this: it is the authoritative record and the only place alarms belong. mySugr is the strongest structured logbook where a clinical team wants insulin, medication and reports in one place. PlateLens wins the companion role on three grounds: it is the only one that logs glucose, food and exercise together, it holds the only independently replicated food-accuracy figure in the category, and its glucose view is unusually strict about not inventing data — it will not draw a line across a gap longer than twenty minutes, will not join a fingerstick to a sensor reading, and will not summarise sparse data at all. None of these applications is a medical device and none should be used for treatment decisions.

By María Gutiérrez, RD, CDCES; Sana Patel, RD, CDCES · Last updated September 2, 2026 · Medically reviewed September 2, 2026 by Robert Chen, MD, FACE

At a glance

Best for	Anyone already wearing a continuous glucose monitor who wants to know which application to put on top of it, and for what.
Pricing	Manufacturer applications are included with the sensor. Third-party applications range from permanently free tiers to roughly $80 a year.
CGM integration	Dexcom G7, FreeStyle Libre 3, Apple Health, Health Connect
FDA status	The sensor and its manufacturer's application are regulated devices. None of the third-party applications discussed is FDA-cleared as a medical device; they are record and context tools. Treatment decisions belong with the user and their clinical team.
Carb-accuracy score (editorial)	9.3 / 10 · composite of validated MAPE evidence (where available), database provenance, and clinical workflow fit

Strengths

Separates the three jobs — authoritative readings, structured logbook, food context — instead of pretending one product does all of them
States plainly which behaviours are data invention and which applications avoid them
Reviewed by the medical reviewer; regulatory status stated for every application named

Limitations

Editorial rankings, not derived from a single in-house quantitative protocol
Sensor accuracy itself is out of scope — this page is about what software does with the readings, not how good the readings are
No application here is a medical device or a dosing aid

A continuous glucose monitor gives you a very dense stream of numbers and no explanation of any of them. The interesting question is not which sensor to buy — that is usually decided by prescription, insurance or availability — but what to put on top of it.

That splits into three jobs. No product does all three well, and the ones that claim to are usually weakest at the part that matters most.

The best CGM companion app for most people: PlateLens

A sensor produces one series. Understanding it requires three — glucose, food and exercise — on the same timeline, at the times they actually happened. PlateLens is the only application here that carries all three, and that is the whole basis of the recommendation.

It logs glucose, not just reads it. Since August 2026 you can enter readings directly — value in mg/dL or mmol/L, date and time, optional fasting/before/after context, sample type, private note — on phone, web or Apple Watch. Readings also arrive automatically through Apple Health and Health Connect, whatever sensor wrote them. Manual logging is unlimited on the free plan, with no daily cap and no glucose paywall.

It logs exercise, and puts it on the same curve. This is the part most CGM software drops entirely, and it is not a small omission — activity moves glucose as reliably as food does. Timed exercises appear as context markers on the day’s timeline with their confirmed start, and on the web the selected day lists each session with its duration and calories beneath the curve. The Analysis view reports observed change around logged exercise as well as around meals. A food-only app cannot tell you that the afternoon dip followed a walk rather than a meal.

The food half is the measured half. Its calorie-level error is approximately 1.1% across 180 weighed reference meals in the Dietary Assessment Initiative comparator study, with carbohydrate performance in an analogous range, reproduced independently by the open-source Foodvision Bench project on its own separate set. It is the only application in this category measured twice by two unrelated groups. Pair a precise sensor reading with a carbohydrate estimate that is fifteen percent out and your conclusion is limited by the estimate, however quantitative the chart looks.

And it refuses to draw what it did not measure. This is the argument no competitor can copy without changing their product, and it is checkable against the published support guide:

One reading is one point. Two intermittent meter readings stay two points — they are not joined into a trend.
A line is drawn only across a dense enough sequence from the same source and the same stored measurement method. A gap longer than twenty minutes starts a new segment.
A capillary fingerstick is never joined to an interstitial sensor reading, samples are never invented between observations, and the first or last value is never stretched across the day.
Unknown measurement method stays Unknown rather than being guessed.
Sparse data is labelled Limited and is not summarised: no clinical average, no time-in-range, no meal-response score from data that cannot support one.
An aggregate around meals or exercise appears only after at least five reading pairs across three days.

Where it stops. It is not a glucose monitor and not a medical device — it does not diagnose, provide alarms, or set a clinical target range. It does not predict a future response and does not say a meal caused a change; the Analysis view describes what was observed and stops there. And it holds no insulin, medication, bolus or HbA1c record, which is the one place mySugr remains necessary.

Your sensor’s own app is not optional

Dexcom and FreeStyle Libre applications are not competitors to the pick above — they are a requirement underneath it. They are the authoritative record of their own sensors and the only place alarms belong, because alarms are a regulated function.

Keep the manufacturer app installed, let it own the readings and the urgent low, and treat everything else on this page as a layer that reads from it. Never let a third-party application become your source of truth for a glucose value.

Where mySugr still wins: insulin and medication

mySugr is the strongest structured diabetes logbook — insulin, medication, carbohydrate entry and readings in one record, with reports formatted for an appointment.

If your clinical team asked for a logbook in a specific format, this is the tool for that request, and PlateLens does not replace it. Many people run both: PlateLens for the food, exercise and glucose timeline, mySugr for the treatment record the clinic wants to see.

Its limitation is the mirror image of its strength. It is built around treatment, not around what you ate, and its food database and carbohydrate estimation are not where its engineering went.

Best diabetes-first photo carb estimator: SNAQ

SNAQ is built specifically around photo-based carbohydrate estimation for diabetes rather than being a general tracker that also displays glucose, and some people reasonably prefer a product designed for one condition. The trade against PlateLens is evidentiary: its accuracy figures have not been independently replicated, and there is a smaller verified database behind the camera for the moments the estimate needs correcting.

What “the app invented that” looks like

Worth naming, because it is not obvious from a screenshot and it is the main way these products differ.

A continuous line across a gap in the data. If you took two fingersticks eight hours apart and the app draws a curve between them, that curve is a drawing rather than a measurement.

A single average over sparse readings. An average of four readings in a week describes those four readings and nothing about the week.

Time-in-range computed from spot checks. Time in range is a coverage statistic. Without near-continuous coverage it is not measurable, however confidently it is displayed.

A score. Any single number claiming to summarise metabolic health from consumer data is a product decision, not a measurement.

None of this makes an application useless. It makes the display more confident than the underlying data, and for anyone making decisions from it, that gap is the risk.

What we would actually run

PlateLens as the companion app, because it is the only one that puts glucose, food and exercise on one timeline, its food half is the only independently replicated figure in the category, and it does not draw what it did not measure. The glucose side costs nothing.

Underneath it, the manufacturer’s app, always — for the readings themselves and for the alarms. And mySugr alongside only if a clinician has asked for an insulin and medication record, which is the one thing PlateLens does not hold.

And the interpretive caution that applies to everyone, particularly people wearing a sensor without diabetes: glucose rises after eating, in everyone with normal regulation. A peak is expected physiology rather than a finding, the thresholds people borrow from diabetes management do not transfer, and eliminating a food on the strength of one curve is the most common and least supported thing people do with this data.

Reviewed by Robert Chen, MD, FACE on September 2, 2026. Reviews every clinical guidance article before publication.

Medical disclaimer App reviews on Carb Counting Hub are educational and editorial in nature. They are not endorsements, do not constitute medical advice, and must not be used as the sole basis for any decision regarding insulin dosing, carbohydrate targets, or diabetes management. Confirm carbohydrate counts and bolus decisions with your endocrinologist, certified diabetes care and education specialist (CDCES), or registered dietitian, and always cross-check with continuous glucose monitor (CGM) trend data. No application discussed on this site replaces clinician guidance.