GPTZero launches 4o detector, claims it beats Pangram on third-party benchmarks
JeremyNguyenPhD · x · 2026-09-25
GPTZero released GPTZero 4o, billing it as the world's most accurate AI text detector: fewer than 1 in 10,000 human texts falsely flagged, and 98.4% detection of Claude Fable. The team claims wins over Pangram on two third-party benchmarks (DetectRL, Epoch AI bench) and says it found major biases in Pangram. One commenter asks the community to verify.
Related event: GPTZero Launches 4o AI Text Detector with Ultra-Low False Positive Rate(2 posts)→
More from Models
- AutoResearchExam: 29-task benchmark ranks agents on 24-hour open-ended ML research — AlexGDimakis · 2026-09-25
- Vision model tier list updated with Opus 5.5, GPT-6 Sol/Luna, and Grok 4.7 — ducha_aiki · 2026-09-25
- Anthropic Accused of Quietly Nerfing Models Weeks After Launch, Opus 5.5 Expected to Follow — iannuttall · 2026-09-25
- TypeSafe AI launches Jev, a 'System One' model for bounded decisions in agent runs — hardimanjames · 2026-09-25
- Qwen Flash Next IQ4_XS beats 27B FP8 on MMLU-Pro, GPQA and GSM8K in community eval — smallDeltaBigEffect · 2026-09-25
- Theo misses the 'vibe' of old Claude 'Fable': Opus is better, but something feels off — teortaxesTex · 2026-09-25