ai-humanizer/ai-humanizer-benchmark-2026

AI Humanizer Benchmark 2026 (Rephrasy)

Every detector result behind the Best AI Humanizer 2026 ranking, one row per (tool, text, detector, run). Nine humanizers, two detectors, four test batches from December 2025 to September 2026.

Nothing here is aggregated into a number you cannot trace. Each row has a source column: a public blog post with screenshots, or a screenshot in the org-card assets folder.

Files

• benchmark.csv – 30 rows. Columns: test_id, date, tool, tool_model, source_text_id, words_out, detector, detector_result_pct, detector_verdict, passed, notes, source
• source_texts.md – the 100-word source paragraph used in the September free-tier runs, and every humanized output that was scored, verbatim.

Test batches

test_id	date	what
seven-tools-dec25	2025-12-19	One AI-written article through 7 tools, scored on ZeroGPT and GPTZero
v3-release-jan26	2026-01-30	100 essays, 50 topics, 600+ runs: v2 vs v3 first-pass rate on GPTZero
v4-release-aug26	2026-08-10	10 fresh documents (250–1,000 words) through v4, GPTZero and ZeroGPT
rerun-loop-sep26	2026-09-03	One 500-word Claude essay, three v4 passes, GPTZero each time
free-tiers-sep26	2026-09-18	100-word paragraph through two free tools, 2 runs each, ZeroGPT

Caveats

• Rephrasy ran every batch. That is why every row carries a source you can open.
• The seven-tool test uses one document per tool. Treat five-point gaps between tools as noise.
• Detectors change monthly. A row describes the detector version of its date.
• GPTZero was not part of the September free-tier batch. Those cells are empty rather than passes.

Rerun it

The source paragraph and every output are in source_texts.md. Paste them into ZeroGPT or GPTZero yourself. If you get different numbers, open a discussion and we will add your rows.

License

CC BY 4.0. Cite as: Rephrasy, AI Humanizer Benchmark 2026, Hugging Face, September 2026.