Yes. ClaudeDetect reads writing style, so it applies to text from any current Claude model, and Veread detects 98.4% of fully AI-written Anthropic text. It does not tell you which version wrote it. Status as of 2026-11-10.
Current Claude versions
- Fable 5.2 and Mythos 5.2. Released 2026-11-09 as the same underlying model. Fable has cyber and biology safeguards and is offered to all users; Mythos is for approved cyber and biomedical organisations. Fable is available to Pro, Max, Ultra, Team and Enterprise users in the Claude apps.
- Opus 5.5. Released 2026-09-22. Anthropic's documentation recommends it as the default starting point.
- Sonnet 5.5. Released 2026-09-28. The mid-priced model of the lineup.
- Haiku 5.5. Released 2026-10-07. The lowest-cost model.
Older models are still active in the API, including Opus 4.5 to 4.8, Sonnet 4.6 and Haiku 4.5. Sonnet 4.5 is deprecated. Anthropic lists a text watermark for these models, which only approved organisations can verify; see the watermark status.
How Veread performs on Claude text
Veread, Fileverity's text model, runs this checker. Fully AI-written text from the Anthropic family is detected 98.4% of the time, while across every kind of AI-written and AI-assisted text the rate is 95.5%. The facts page lists the sources. Texts that Claude only polished or partly rewrote are harder than fully generated ones.
How to check text from each version
- Copy the text into the box on the Claude AI detector, with at least 100 words.
- Read the verdict and the highlighted passages first, then the family similarity.
- A high Claude similarity means the style resembles Claude. It does not prove which model or person wrote the text.
- Texts shorter than 100 words carry too little signal; the facts page lists the minimum.
Sources: Anthropic, Claude Fable and Mythos 5.2; Claude API models overview; model deprecations.