Claude Opus Glossary

The high-capability tier of Anthropic's Claude model family and the source of the dependency described in Opus Addict.

Claude Opus is the high-capability tier in Anthropic’s Claude model family. Anthropic positions it for difficult reasoning, writing, coding, and agentic work, usually at a higher price and latency than the Sonnet tier. The exact benchmark lead and price ratio change by model generation.

Claude Opus was the principal drafting model for much of this Dictionary’s early development. Opus Addict names the resulting dependence on a closed frontier tier and the operational risks that remain even when the model performs well.

A 16-day cost review in April 2026 found that cache writes accounted for 72.6% of the operator’s Anthropic spending, while output tokens accounted for 4.8%. That result helped motivate the local-compute architecture described in Sovereign Compute.

See also

• Claude — the model family
• Opus Addict — the structural-dependency entry
• Anthropic
• Claude Sonnet, Claude Haiku — peer tiers
• Sovereign Compute

Return to Dictionary All Entries (A–Z) For Students Other Writing Capstone 2.0