Many people have moved from ChatGPT to Claude or Gemini for writing, partly because the output feels more natural. So does GPTZero still catch it?
Quick Answer
GPTZero says it detects Claude, ChatGPT and Gemini, and in practice raw output from any of them can be flagged. A more natural style doesn't make text undetectable, because detectors score statistical patterns, not how pleasant the writing sounds. Read the full GPTZero guide for how it scores text.
Why a Natural Style Isn't the Same as Undetectable
Claude's prose often reads less stiff than early ChatGPT output, and that is why people assume detectors will miss it. But GPTZero measures things like how predictable the word choices are and how much sentence rhythm varies. A model can sound good to a human reader and still be statistically very regular.
How the Models Compare
These are general characteristics, not our measurements.
Detectors retrain as new model versions appear, so any claim that one model is "invisible" tends to have a short shelf life.
What Actually Changes the Score
Usually not enough
- −Switching to a different model
- −Swapping synonyms
- −Prompting "write like a human"
- −Light manual edits
Helps more
- +Restructuring sentences
- +Varying length and rhythm
- +Adding your own specifics
- +Humanizing, then proofreading
A Simple Workflow for Claude or Gemini Drafts
Draft with the model you prefer
Focus on content and structure first.
Humanize the draft
Paste it into Fresh Text to rebuild rhythm and structure instead of swapping words.
Add what only you know
Examples, sources and opinions that come from you.
Proofread and follow the rules
Check facts and tone, and follow the AI policy that applies to you.
Frequently Asked Questions
Does GPTZero detect Claude?▾
GPTZero says its detector covers several major models, including Claude, ChatGPT and Gemini. Detection is statistical, so results on any one text can vary.
Is Claude harder to detect than ChatGPT?▾
It depends on the text and the model version. Detectors are tuned on data from popular models and update regularly, so no model is reliably undetectable. Claude's writing often has a different style, but detectors look at patterns, not brand.
Does Turnitin detect Claude?▾
Turnitin's AI detector is designed to recognize text from large language models in general, not only ChatGPT. Results vary with length and editing.
Does the model I use change my AI score?▾
Somewhat, because models write with different habits, but what moves the score most is how much of the text is raw, how long it is and whether its structure was rewritten.
How do I lower an AI score on Claude text?▾
The same way as for any AI text: restructure sentences, vary rhythm, add your own specifics and run the draft through a humanizer before a final proofread.
Conclusion
GPTZero can flag Claude as well as ChatGPT or Gemini. Which model you use matters less than how much of the text is raw and how deeply you rework it.