Grok now watermarks generated text — users weigh how to avoid it
walkingriver · x · 2026-11-01
A developer asks what strategies exist now that Grok watermarks its generated text, beyond running local models — or whether it even matters. He notes he doesn't care for code output. The thread touches on watermarking mechanisms across vendors, with local deployment remaining the most complete workaround.
More from Models
- Power user: Astra is smart but lazy, gives up too easily in real tasks — ChrisGPT · 2026-11-01
- What tok/s would you tolerate running local LLMs? 15 is the cutoff — challis88ocarina · 2026-11-01
- Dev rants: Claude 'cannot comprehend' word choice and phrasing, 'it cannot write' — hewarsaber · 2026-11-01
- How different AIs "chose" a random number: 7 and 47 everywhere — gabrieleremita · 2026-11-01
- Newest LLM isn't automatically best for business: reliability and cost beat benchmarks — DavidLinthicum · 2026-11-01
- Using AI to crack a spatial folding puzzle on a Punjabi TV quiz show — satnam6502 · 2026-11-01