How do I make a ROC curve?

Use the generator above: enter your operating points as false positive rate (x) and true positive rate (y), each between 0 and 1. The curve, the random-classifier diagonal, and the AUC render instantly, and you can add more curves to compare models. Download it as a scalable SVG — free, with no sign-up.

What is AUC and how is it calculated?

AUC is the Area Under the ROC Curve, a value between 0 and 1 that summarizes performance across all thresholds. It equals the probability that the model ranks a random positive above a random negative. This tool computes it with the trapezoidal rule: it sums the areas of the trapezoids between consecutive (FPR, TPR) points after sorting them by FPR.

What is a good AUC value?

AUC = 0.5 is no better than random guessing. As a rough guide, 0.7–0.8 is acceptable, 0.8–0.9 is good, and above 0.9 is excellent. An AUC below 0.5 means the model is worse than chance, which usually indicates inverted predictions. Because AUC is threshold-independent, it is best used to compare models rather than to choose a deployment threshold.

What is the difference between TPR and FPR?

The true positive rate (TPR = TP / (TP + FN), also called sensitivity or recall) is the share of real positives the model catches. The false positive rate (FPR = FP / (FP + TN), equal to 1 − specificity) is the share of real negatives it wrongly flags. A ROC curve plots TPR on the y-axis against FPR on the x-axis.

Can I compare several models on one ROC chart?

Yes. Click "Add curve" to plot a second or third model on the same axes. Each curve gets its own color and its AUC in the legend, so you can see at a glance which model separates the classes better across thresholds.

Can I export the ROC curve for a paper or report?

Yes. Download it as an SVG — a vector format that stays sharp at any size and drops straight into a manuscript, thesis, poster, slide deck, or Jupyter notebook, or opens in Illustrator or Inkscape for final edits. Every version is free to download.

Data Science Tool

ROC Curve Generator with AUC

Free ROC curve generator — plot true positive rate against false positive rate, compare models on one chart, and compute AUC. Enter your points, export SVG.

Plot TPR vs FPR — exact SVG with AUCCompare several models on one chartTrapezoidal AUC computed as you typeDownload a publication-ready SVG — free

Enter your TPR / FPR operating points — renders an exact ROC curve with AUC, free

Title

AUC 0.868

FPR (x)TPR (y)

Enter operating points as false positive rate (x) and true positive rate (y), each between 0 and 1. The (0,0) and (1,1) corners and the AUC are added automatically.

ROC Curve Examples

Exact engine renders — single models, comparisons, and diagnostic tests

View:

Single Classifier

Exact engine render — one classifier with its AUC and the random baseline.

aucsensitivityspecificity

Model Comparison

Exact engine render — several models overlaid, each with its AUC in the legend.

comparisonmulti-modelauc

Excellent Classifier

Exact engine render — a strong model whose curve hugs the top-left corner.

high-aucexcellenttop-left

Weak Classifier

Exact engine render — a weak model whose curve sits near the random diagonal.

low-aucnear-diagonalweak

Medical Diagnostic Test

Exact engine render — a diagnostic test framed as sensitivity vs 1 − specificity.

medicaldiagnosticsensitivity

Three-Model Comparison

Exact engine render — three models with different AUCs on one chart.

three-modelthresholdsauc

What is a ROC curve?

A ROC (Receiver Operating Characteristic) curve shows how well a binary classifier separates two classes as you vary its decision threshold. It plots the true positive rate (sensitivity, or recall) on the y-axis against the false positive rate (1 − specificity) on the x-axis. Each point is the model at one threshold: lowering the threshold catches more true positives but also more false positives, tracing a curve from the bottom-left (0, 0) to the top-right (1, 1). A perfect classifier hugs the top-left corner; a useless one follows the diagonal. This generator draws the curve and computes the area under it (AUC) from the operating points you enter, and it can overlay several models on one chart to compare them.

True positive rate, false positive rate, and thresholds

True Positive Rate (TPR) = TP / (TP + FN) — also called sensitivity or recall; the share of real positives the model catches.
False Positive Rate (FPR) = FP / (FP + TN) — equal to 1 − specificity; the share of real negatives wrongly flagged.
Each threshold gives one (FPR, TPR) point; sweeping the threshold from strict to lenient traces the ROC curve.
The dashed diagonal is the random classifier (AUC = 0.5); a good model bows toward the top-left corner.

What the AUC means

AUC is the Area Under the ROC Curve, a single number between 0 and 1 that summarizes performance across all thresholds. It equals the probability that the model ranks a random positive case above a random negative one. AUC = 0.5 is no better than chance, 0.7–0.8 is generally considered acceptable, 0.8–0.9 is good, and above 0.9 is excellent; below 0.5 means the predictions are inverted. Because AUC is threshold-independent, it is useful for comparing models, and this tool computes it with the trapezoidal rule from the points you provide. AUC does not tell you the best threshold to deploy — for that, look at the curve itself and the cost of false positives versus false negatives.

How to make your ROC curve

Enter each operating point as a false positive rate (x) and a true positive rate (y), both between 0 and 1.
The (0, 0) and (1, 1) corners and the AUC are added and computed for you as you type.
Add a second or third curve to compare models on the same axes — each gets its own color and AUC in the legend.
Download a clean, scalable SVG for your paper, notebook, or slides — free, with no sign-up.

ROC curve vs precision-recall curve

ROC curves are the standard way to report classifier performance, but on heavily imbalanced data they can look overly optimistic because the false positive rate stays low when negatives vastly outnumber positives. In that situation a precision-recall curve is often more informative, since precision reacts directly to false positives among the predicted positives. A good workflow is to read the ROC curve and AUC alongside the raw counts in a confusion matrix so you can see the trade-off in absolute terms, then pick the threshold that matches the cost of each error for your problem.

Frequently Asked Questions

Try it free

View All Free Tools

ROC Curve Generator with AUC

Free ROC curve generator — plot true positive rate against false positive rate, compare models on one chart, and compute AUC. Enter your points, export SVG.

Plot TPR vs FPR — exact SVG with AUCCompare several models on one chartTrapezoidal AUC computed as you typeDownload a publication-ready SVG — free

What is a ROC curve?

True positive rate, false positive rate, and thresholds

True Positive Rate (TPR) = TP / (TP + FN) — also called sensitivity or recall; the share of real positives the model catches.

False Positive Rate (FPR) = FP / (FP + TN) — equal to 1 − specificity; the share of real negatives wrongly flagged.

Each threshold gives one (FPR, TPR) point; sweeping the threshold from strict to lenient traces the ROC curve.

The dashed diagonal is the random classifier (AUC = 0.5); a good model bows toward the top-left corner.

What the AUC means

How to make your ROC curve

Enter each operating point as a false positive rate (x) and a true positive rate (y), both between 0 and 1.

The (0, 0) and (1, 1) corners and the AUC are added and computed for you as you type.

Add a second or third curve to compare models on the same axes — each gets its own color and AUC in the legend.

Download a clean, scalable SVG for your paper, notebook, or slides — free, with no sign-up.

ROC curve vs precision-recall curve