Skip to content

Head to head · Detector v10 · October 2026

ParaTrace against Pangram

21 results on the same public tests, with Pangram’s figures as Pangram published them and ours measured on the same texts.

ParaTrace catches 100 % of fully AI-written student essays, the same as Pangram, and misses fewer AI texts than both Pangram 4 and Pangram 3 in a test of writing by well-known authors. Across the 21 results measured on the same tests, it is within 5 points of Pangram 4 on 17 and of Pangram 3 on 18. The remaining gaps are on humanized text, above all under 50 words, and on very short texts, where it is just over 5 points behind Pangram 4.

Within 5 points

  • Against Pangram 417 of 21 results
  • Against Pangram 318 of 21 results

Results where ParaTrace is level, ahead, or no more than 5 points behind.

Strengths

Where ParaTrace matches or beats Pangram

Ahead on well-known authors

0.84 %

of AI texts missed in a test of writing by well-known authors, against 2.86 % for Pangram 4 and 5.05 % for Pangram 3.

Level on student essays

1.000

AUROC on all 4 student-essay tests, the same as Pangram 4, with 100 % of fully AI-written essays caught.

Ahead of Pangram 3 on short texts

94.66 %

of AI texts under 50 words caught at 1 % false alarms, against 88.87 % for Pangram 3.

The breakdown

All 21 results, test by test

Compare with
Student essaysStudent essays set against essays written entirely by AI, AI essays disguised by a paraphrasing tool, and student essays lightly polished by AI.
ResultPangram 4ParaTraceGap
Fully AI-writtenAUROC1.0001.000Level
Fully AI-writtenAI caught at 1 % false alarms100 %100 %Level
AI-paraphrasedAUROC1.0001.000Level
AI-paraphrasedAI caught at 1 % false alarms100 %98.9 %−1.1
Polished by AIAUROC1.0001.000Level
Polished by AIAI caught at 1 % false alarms100 %98.9 %−1.1
OverallAUROC1.0001.000Level
OverallAI caught at 1 % false alarms100 %99.4 %−0.6
Eight kinds of writingBooks, news, scientific abstracts, recipes, forum posts, reviews, encyclopedia articles and poetry, set against untouched text from recent AI chat models.
ResultPangram 4ParaTraceGap
All textsAUROC1.00000.9999−0.01
All textsAI caught at 1 % false alarms100 %99.94 %−0.06
Novels, news, reviews and moreNews, blogs, product and restaurant reviews, novel excerpts and résumés, with AI versions at full length, under 50 words, and after a humanizer tool.
ResultPangram 4ParaTraceGap
Full lengthAUROC1.00001.0000Level
Full lengthAI caught at 1 % false alarms100 %99.91 %−0.09
Under 50 wordsAUROC0.99990.9965−0.34
Under 50 wordsAI caught at 1 % false alarms99.70 %94.66 %More than 5 points behind, −5.04
Humanized, full lengthAUROC0.99960.9959−0.37
Humanized, full lengthAI caught at 1 % false alarms98.93 %92.03 %More than 5 points behind, −6.90
Humanized, under 50 wordsAUROC0.98100.8778More than 5 points behind, −10.32
Humanized, under 50 wordsAI caught at 1 % false alarms73.32 %36.43 %More than 5 points behind, −36.89
Human writingHuman texts wrongly flagged0.00 %0.10 %−0.10
Well-known authorsBlog posts, fiction and scientific writing by well-known authors, set against AI passages from a simple prompt or written to copy each author’s style.
ResultPangram 4ParaTraceGap
AI textsAI texts missed2.86 %0.84 %ParaTrace ahead, +2.02
Human textsHuman texts wrongly flagged0.00 %1.62 %−1.62

The gap is in points, ParaTrace’s figure against Pangram’s, so a minus means ParaTrace is behind. ▲ marks where ParaTrace is ahead, and ▼ where it is more than 5 points behind. For AUROC, one point is 0.01.

How we compared

Like for like

Only matching results.
Every Pangram figure here is one Pangram published for a public test, in its Pangram 4 technical report of July 2026. We scored ParaTrace on the same texts and measured it the same way, and left out every result we could not match exactly.
AI caught at 1 % false alarms.
Each detector is set so that 1 in 100 of the test’s human texts is flagged, and we count the share of AI texts it then catches.
AUROC.
How well a detector ranks AI text above human text across every possible setting, from 0.5 for a coin toss to 1 for perfect.
Missed and wrongly flagged.
Each detector at its standard setting. For ParaTrace that is its Strict setting, applied to each text as a whole.
Points.
Gaps are in percentage points, and a minus always means ParaTrace is behind, even where a lower figure is better. “Level” means the same figure at the precision Pangram published.
Versions.
ParaTrace is detector v10, the model this site runs, scored in October 2026 with the service’s model and scoring code on our own hardware. Both Pangram versions are as Pangram reports them. For the test on well-known authors, Pangram 3 is version 3.3.2.
About Pangram.
Pangram is a product of Pangram Labs, Inc., which has not reviewed this comparison.
More results.
Our benchmark report covers many more models, evasion attacks and kinds of writing, and a field test on scientific papers repeats the check NeurIPS ran with Pangram.