November 2026 · Independent humanizer benchmarks. Explore the evidence

AI humanizers for essays

Academic and application essay categories together. Category scores are taken from the HumanizeBench v2 per-category output, before entry-level penalties.

Measurement coverage5,252 / 5,313
Current scores

Scores are provisional while 58 detector checks remain unresolved. Rankings update as measurements arrive. Provider exclusions: 3. Excluded checks have no numeric score. Readability and meaning: 759/759 recorded.

Inspect coverage and recorded evidence →
Rank ⓘHumanizerCategory average ⓘ ↓Bypass ⓘMeaning ⓘReadability ⓘPenalties ⓘCoverageEvidence
133 examples
Rephrasyrephrasy.ai
60.5078.035.773.4△ −12.0
231/231
233 examples
CleverHumanizercleverhumanizer.ai
60.4077.325.475.8△ −10.0
231/231
333 examples
UndetectedGPTundetectedgpt.ai
57.5061.346.879.8△ −10.0
231/231
433 examples
Superhumanizersuperhumanizer.ai
53.8046.543.274.6△ −10.0
231/231
533 examples
StealthGPTstealthgpt.ai
51.5538.646.479.9△ −12.0
231/231
633 examples
HIX Bypasshixbypass.com
50.7545.441.376.9△ −20.0
231/231
733 examples
Qwen 2.5 14B HumanizerVerified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.arunsingh80475/qwen2.5-14b-humanizer-v4-lora
47.6566.221.178.1△ −15.0
230/231
833 examples
SmartHumanizersmarthumanizer.ai
46.3029.050.277.6△ −10.0
231/231
933 examples
GPTinfgptinf.com
45.3037.741.471.7△ −10.0
231/231
1033 examples
Humanizer 12B · OptiQ 4-bitVerified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.mlx-community/humanizer-12b-OptiQ-4bit
44.2536.360.977.1△ −7.0
231/231
1133 examples
Smaug HumanizerVerified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.abacusai/Smaug-Humanizer
42.6518.576.581.3△ −1.0
231/231
1233 examples
Humanizer 12B · GGUFVerified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.jialinyyzz/humanizer-GGUF
41.8525.569.680.0△ −2.0
231/231
1333 examples
Humanizer 12BVerified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.jialinyyzz/humanizer
41.8525.569.680.0△ −2.0
231/231
1433 examples
Open Humanizer 4BVerified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.akhaliq/open-humanizer-4b
41.050.096.685.1△ −10.0
231/231
1533 examples
Humanizer 12B · MLX 8-bit (cowWhySo)Verified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.cowWhySo/humanizer-12b-mlx-8bit
41.0025.366.579.0△ −4.0
230/231
1633 examples
Humanizer 1B · OptiQ 4-bitVerified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.mlx-community/humanizer-1B-OptiQ-4bit
39.900.677.171.2△ −17.0
229/231
1733 examples
Humanizer 12B · MLX 8-bit (npario)Verified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.npario/humanizer-12b-mlx-8bit
39.4524.665.978.8△ −5.0
229/231
1833 examples
Humanizer · CMFVerified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.infosave/humanizer-cmf
39.4030.265.978.0△ −3.0
231/231
1933 examples
HumanizerAI · GGUFVerified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.mradermacher/humanizerAI-GGUF
38.951.075.982.6△ −24.0
199/231
2033 examples
GoHumanizeVerified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.gohumanize/gohumanize-open-humanizer
38.850.292.085.1△ −7.0
231/231
2133 examples
Rehumanizerehumanize.io
30.200.860.282.2△ −6.0
231/231
2233 examples
Undetectable AIundetectable.ai
26.350.555.679.6△ −18.0
231/231
2333 examples
Humanizer 3BVerified by HumanizeBenchWe ran this model ourselves on the benchmark source examples and checked its execution records. Verification confirms the tested model’s provenance; it does not guarantee performance or imply that every detector returned a result.mondk/Humanizer-3B-Safetensors
19.251.312.372.1△ −20.0
208/231

The selected categories’ original scorer values are averaged, before entry-level penalties. Scores use available detector measurements. Coverage and methodology are available above.

For this view, the category column is the mean of the two recorded category scores. It is not the overall composite.