November 2026 · Independent humanizer benchmarks. Explore the evidence

Meaning-preservation results

Semantic preservation, without using detector scores. All values come from our shared evaluation.

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 →
EntryRaw cosineRescaled meaningReadabilityEvidence
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
0.991696.685.1Read outputs →
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
0.980192.085.1Read outputs →
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
0.941476.581.3Read outputs →
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
0.938275.982.6Read outputs →
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
0.924277.171.2Read outputs →
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
0.924169.680.0Read outputs →
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
0.924169.680.0Read outputs →
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
0.916366.579.0Read outputs →
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
0.914765.978.0Read outputs →
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
0.914765.978.8Read outputs →
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
0.902260.977.1Read outputs →
Rehumanizerehumanize.io
0.898760.282.2Read outputs →
Undetectable AIundetectable.ai
0.885655.679.6Read outputs →
SmartHumanizersmarthumanizer.ai
0.875550.277.6Read outputs →
UndetectedGPTundetectedgpt.ai
0.866146.879.8Read outputs →
StealthGPTstealthgpt.ai
0.864446.479.9Read outputs →
HIX Bypasshixbypass.com
0.852241.376.9Read outputs →
Superhumanizersuperhumanizer.ai
0.852243.274.6Read outputs →
GPTinfgptinf.com
0.851841.471.7Read outputs →
Rephrasyrephrasy.ai
0.832835.773.4Read outputs →
CleverHumanizercleverhumanizer.ai
0.806125.475.8Read outputs →
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
0.789221.178.1Read outputs →
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
0.758012.372.1Read outputs →