Benchmarks

Academi.cx vs ZeroGPT

We benchmarked Academi.cx and ZeroGPT against Turnitin across 429 documents, measuring score differences, false flags, and missed AI.

Brayden · 2 min read

For this benchmark, we submitted the same 429 documents to Academi.cx and ZeroGPT through their APIs, then compared each tool’s scores with the documents’ actual Turnitin AI scores. Together those documents add up to close to 944,000 words, averaging about 2,200 words each, ranging from a 311 word short answer to a 25,000 word thesis.

How close were the predictions

Across all 429 documents, Academi.cx averaged 1.92 points off the real Turnitin score. ZeroGPT averaged 22.23 points off, the widest gap of any tool covered in this benchmark so far. Its median error sits at 14.4, meaning even the typical prediction, not just the worst ones, missed by double digits.

Paired scatterplots show the AI score for all 429 documents against Turnitin. Academi.cx appears in blue and ZeroGPT in gray, with identical 0 to 100 percent axes. The dashed diagonal marks matching scores. Mean absolute difference from Turnitin is 1.92 percentage points for Academi.cx and 22.23 for ZeroGPT. No documents are omitted; some points overlap.

Breaking it down by document type

The 429 documents split into 220 that Turnitin scored as 0% AI, 99 it scored as 100% AI, and 110 that fell somewhere in between.

Grouped bars show average absolute deviation from Turnitin in percentage points. Across all 429 documents, Academi.cx averages 1.92 points and ZeroGPT 22.23. For 220 documents Turnitin scored 0%, the averages are 0.18 and 11.87; for 110 documents scored 20 to 99%, 3.95 and 27.08; for 99 documents scored 100%, 3.53 and 39.86. Blue represents Academi.cx and gray ZeroGPT. Lower is better, and all bars share the same scale.

On the 0% documents, Academi.cx averaged 0.18 points of error, ZeroGPT averaged 11.87. On the 100% documents the gap widens further, Academi.cx averaged 3.53 points off versus ZeroGPT at 39.86, its single weakest number in the set, meaning it consistently undershot genuinely AI written documents by a wide margin. The mixed bucket also ran hot for ZeroGPT at 27.08 points of error compared to Academi.cx's 3.95.

False flags and missed AI

Academi.cx had a 0% false flag rate and a 0% missed AI rate across all 429 documents, meaning it never scored a genuinely 0% document above 20%, and never scored a genuinely 100% document below 80%. ZeroGPT came in at a 20.5% false flag rate and a 66.7% missed AI rate, meaning two out of every three fully AI written documents in this set scored below 80% AI on ZeroGPT.

Sentence level accuracy

We also checked how often each tool correctly classified individual sentences as AI written or human written, matched against Turnitin's sentence level highlights. Across all 429 documents, Academi.cx correctly classified 96.64% of sentences. ZeroGPT came in at 71.69%.

The takeaway

Across all 429 documents, Academi.cx closely matched Turnitin’s scores on every measure checked, while ZeroGPT posted the largest average error and the highest missed AI rate of any tool in this benchmark, particularly on documents Turnitin confirmed were fully AI written.

See the full benchmark results.

Related reading