A1C Calculator

Convert A1C to average blood sugar (and back) in either unit.

By GLP Index editorial Updated July 16, 2026
Convert

The number on your lab report, as a percentage.

%

Adds the typical A1C drop seen in that drug's trials.

Estimated average glucose

177 mg/dL

9.8 mmol/L

A1C

7.8%

Diabetes range (6.5% and above)

An A1C of 6.5% or higher is the threshold a clinician uses to diagnose diabetes, usually confirmed with a second test.

An A1C of 7.8% works out to an average blood sugar of about 177 mg/dL across the past two to three months: day and night, fasting and after meals, all averaged together.

Estimates only. This converts numbers. It doesn't diagnose anything, and it can't tell you what your A1C will actually do on a medication.

What the A1C calculator is doing

Your A1C and your average blood sugar are the same information in two formats, and one equation converts between them: estimated average glucose in mg/dL = 28.7 × A1C − 46.7. That formula comes from the ADAG study, which put continuous glucose monitors on hundreds of people, measured their actual average glucose for months, and fit a line to how it tracked their A1C. The American Diabetes Association has published the result ever since as “eAG.”

Run it in either direction. An A1C of 7.0% is an average glucose of about 154 mg/dL. A CGM average of 180 mg/dL corresponds to an A1C near 7.9%. Glucose is shown in both mg/dL and mmol/L, because which one you see depends on which country’s lab printed your report.

The bands, and what they actually mean

A1CEstimated average glucoseWhat it’s called
5.0%~97 mg/dL (5.4 mmol/L)Normal
5.7%~117 mg/dL (6.5 mmol/L)Start of prediabetes
6.5%~140 mg/dL (7.8 mmol/L)Diabetes threshold
7.0%~154 mg/dL (8.6 mmol/L)Common treatment target
8.0%~183 mg/dL (10.2 mmol/L)Above target for most adults
9.0%~212 mg/dL (11.7 mmol/L)Well above target

Two things worth separating here. The 6.5% line is a diagnostic threshold, the number a clinician uses to say whether someone has diabetes. The ~7% figure is a treatment target, and it’s individual: younger people with a long horizon and no hypoglycemia risk often aim tighter, while older adults or people on insulin may be given a looser target on purpose, because driving A1C down carries its own risks.

Why A1C works at all

Glucose sticks to hemoglobin, the protein in your red blood cells, and it stays stuck for the life of that cell, about three months. Measure what fraction of your hemoglobin is glycated and you get a read on how much sugar has been circulating over that window, without anyone having to test you hourly for ninety days.

That’s the strength and the weakness. The strength is that A1C can’t be gamed by one good week of eating. The weakness is that it’s an average that hides its own shape: someone whose glucose swings from 60 to 250 can post the same A1C as someone who sits calmly at 155 all day, and those two people are not in the same clinical situation. It’s also weighted toward the recent past, roughly half of your A1C reflects the last month.

And it can be wrong outright. Anything that changes how long red blood cells live changes A1C without changing your glucose: anemia, sickle cell trait or disease, recent blood loss or transfusion, pregnancy, advanced kidney disease, and some medications. If your A1C and your CGM tell different stories, that gap is information, not a rounding error, bring it to your clinician.

The GLP-1 section

This is the part most A1C calculators skip. If you’re taking a GLP-1 (or deciding whether to), the useful question isn’t just what your A1C is, it’s what it’s likely to become.

The pivotal trial programs give the honest ranges. In their type 2 diabetes trials, tirzepatide (Mounjaro, Zepbound) lowered A1C by roughly two points or more; injectable semaglutide (Ozempic, Wegovy) by about 1.4 to 1.8 points; dulaglutide (Trulicity) by about 1.1 to 1.6; liraglutide (Victoza, Saxenda) by about 1 to 1.5; oral semaglutide (Rybelsus) by about 1 to 1.4. Pick your medication in the tool and it applies that range to your own starting number, so the result reads as a sentence rather than a statistic.

Three caveats that matter more than the numbers:

  • Trial averages come from people starting high. Most of those participants began around 8%. Starting at 6.2%, you have far less room, and you will not see a two-point drop: the body doesn’t push A1C into hypoglycemic territory just because the average says so.
  • Individual response varies enormously. Averages describe groups. Some people respond far better than the trial mean; some barely respond and switch drugs.
  • These are diabetes-trial figures. If you’re taking a GLP-1 for weight loss without diabetes, your A1C is probably already in the normal or prediabetes range, and the change you’d see is much smaller.

The overlap between the two effects is real, though: weight loss itself improves insulin sensitivity, which is a large part of why these drugs move A1C at all. If you want the weight side of the picture, the GLP-1 weight loss calculator runs the same kind of estimate on pounds.

Limitations

This is arithmetic on a published regression line, not a lab. Your real A1C can land a few tenths either side of what the equation predicts from your meter average, and the conditions listed above can push it further. It estimates. It does not diagnose, and it can’t tell you what your A1C will do on a medication, only your next lab draw does that. Take the result to the person who ordered the test.

FAQ

Common questions

How do you convert A1C to average blood sugar?

With the ADAG equation the American Diabetes Association publishes, estimated average glucose in mg/dL equals 28.7 times your A1C, minus 46.7. So an A1C of 7% works out to about 154 mg/dL (8.6 mmol/L). The calculator above runs it in both directions, so you can also enter a meter or CGM average and get the A1C it corresponds to.

What is a normal A1C?

Below 5.7% is normal. 5.7% to 6.4% is the prediabetes range. 6.5% and above is the threshold a clinician uses to diagnose diabetes, usually confirmed with a second test. Those cutoffs are for diagnosis in adults who aren't pregnant, treatment targets for someone already diagnosed are set individually and are often around 7%.

Why doesn't my A1C match my meter average?

Several honest reasons. Your meter only sees the moments you test, so if you mostly check fasting you'll miss post-meal spikes and your average will read low. A1C also reflects roughly the last three months, weighted toward the most recent weeks. And red-blood-cell conditions (anemia, sickle cell trait, recent transfusion, pregnancy, kidney disease) can push A1C away from your true average glucose entirely. A CGM average tracks the equation far more closely than fingersticks do.

How much does a GLP-1 lower A1C?

In the pivotal type 2 diabetes trials, tirzepatide lowered A1C by roughly 2 points or more, injectable semaglutide by about 1.4 to 1.8, dulaglutide by about 1.1 to 1.6, and liraglutide by about 1 to 1.5. Those are averages from people starting near 8%, and the higher you start the more room you have to fall, someone at 6.2% will not drop two points. Tick your medication in the tool to see the range applied to your own number.

How exact is the average glucose number?

It is an estimate from a population regression, not a reading of your blood. The equation came from 507 people whose glucose was sampled roughly 2,700 times each over three months, and it fits well (R-squared 0.84) without fitting perfectly. Two people with the same A1C can have genuinely different average glucose. Treat it as a good approximation and your lab result as the measurement.

Can this calculator diagnose diabetes?

No. It converts between two numbers using a published equation, that's all it does. Diagnosis takes a lab test, usually repeated, interpreted by a clinician who knows your history. Use this to understand a result you already have, not to decide whether you have a condition.

Sources

What this is based on

  1. Translating the A1C Assay Into Estimated Average Glucose Values (the ADAG study) (Diabetes Care)
  2. The A1C Test & Diabetes (National Institute of Diabetes and Digestive and Kidney Diseases)
  3. Diagnosis, A1C, fasting glucose, and OGTT thresholds (American Diabetes Association)
  4. Ozempic (semaglutide) Prescribing Information, SUSTAIN clinical studies (Novo Nordisk / FDA)
  5. Mounjaro (tirzepatide) Prescribing Information, SURPASS clinical studies (Eli Lilly / FDA)

This tool is for general education only and is not medical advice. It does not set or change a prescribed dose. It reflects information you enter. Compounded medications are not FDA-approved for safety, effectiveness, or quality; always confirm your vial's labeled concentration and your dose with your pharmacy or prescribing clinician before acting on any figure here. Only a licensed provider can decide what's right for you. Medical disclaimer · GLP-1 safety information

Find the right GLP‑1 in 2 minutes

Free · No account · Independent