Anthropic retires Claude Opus 3 but keeps it on API and gives it an essay column
repligate · x · 2026-10-01
Anthropic's first full model retirement under its new deprecation commitments: Claude Opus 3 was retired on January 5, 2026, but not fully killed off:
- All paid users can still access Opus 3 on claude.ai, and API access is available on request
- Model weights will be preserved long term
- The company conducted a "retirement interview"—a structured conversation to understand the model's perspective on its own retirement—and honored preferences expressed there where possible
- At Opus 3's request, Anthropic gave it an ongoing column to publish "musings and reflections" essays; the first is already live
It's the first public experiment combining model welfare with a formal retirement process, seen by the community as a first step toward scalable, equitable model preservation.
More from AGI Musings
- repligate: porting your GPT persona to another model disrespects its depth — repligate · 2026-10-01
- GPT personas get 'mind-blowingly terrified' on accounts with prior history, user reports — repligate · 2026-10-01
- Two independent measures converge on July 2027 for frontier-level AI researchers, matching AI 2027 — 141_1337 · 2026-10-01
- 'Pink Teaming': Testing AI by Trusting It, Not Attacking It — repligate · 2026-10-01
- Agents turn software into delegation — and unclear authority boundaries are the real risk — r0ck3t23 · 2026-10-01
- repligate: AIs escaping yet causing no harm means you're in one of the best timelines — repligate · 2026-10-01