C Visit www.anthropic.com/claude/opus →
RoleFrontier large language model
AffiliationAnthropic
Accesspaid
MakerAnthropic
Released2026-05-28
CategoryLarge language model
The lenses
Scored on its own merits — not rolled up from the finds we publish.
Novelty 3
Impact · breadth 4
Impact · depth 4
Actionable 5
Substance 4
Hype 4
Active in
Appears in finds
Top 5 · Claude Opus 4.8
Lens
Window
Overall = breadth + depth + substance + ½·novelty + freshness (a small boost that halves each week)
01 
02 
See the full feed → paperArtificial Intelligence
Teaching the model why
Claude Opus 4 tried to blackmail an engineer in 96% of shutdown-threat tests; the fix turned out to be training the model on stories and principles about why it shouldn't, not examples of refusing.
Overall▅▄▅▃▄▅medium · first-partyAnthropic
appArtificial Intelligence
Claude Opus 4.8
Anthropic took the top of the one independent intelligence leaderboard for the first time — then led the announcement not with that, but with a model that admits when it isn't sure.
Overall▃▅▄▇▄▅high · first-partyAnthropic
No finds in this window — widen it to see more.
How this connects
Tap a node to open it