# Anthropic adds its hidden text watermark to Mythos 5, Opus 4.7, Opus 4.6 and Sonnet 4.6 on November 3

On the two 4.6 models the mark applies only when developers leave sampling parameters at their defaults, so changing one parameter removes it.

- Published: 2026-11-01T06:20:01.798Z
- Canonical: https://polylog.news/ai/2026-11-01/anthropic-adds-its-hidden-text-watermark-to-mythos-5-opus-4
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [Polylog editors](https://polylog.news), [Anthropic News](https://www.anthropic.com/claude-opus-5-5), [The Decoder](https://the-decoder.com/anthropics-watermark-rollout-leaves-a-loophole-on-two-older-claude-models/), [Pasquale Pillitteri](https://pasqualepillitteri.it/en/news/27785/claude-watermark-november-3-opus-4-7-sonnet-4-6), [Pasquale Pillitteri](https://pasqualepillitteri.it/en/news/10649/claude-watermark-remover-github-tested), [AGTP on X](https://x.com/AGTPinsights/status/2116545109651067002)

Anthropic emailed API customers that starting Tuesday, November 3, it will watermark text generated by Claude Mythos 5, Opus 4.7, Opus 4.6 and Sonnet 4.6. Newer models, including Opus 5.5, already carry the same mark. The rollout covers the…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-11-01/anthropic-adds-its-hidden-text-watermark-to-mythos-5-opus-4 (subscription information: https://polylog.news/pricing).