# Claude Opus [[Anthropic]]'s most capable generally available model tier, designed for complex reasoning, long-horizon agentic coding, and high-autonomy work. The Opus line sits between [[Claude Sonnet]] (balanced performance and cost) and the restricted Mythos class ([[Claude Fable 5]] / [[Claude Mythos 5]]) in Anthropic's model hierarchy. The naming convention is poetic: an opus is a long, ambitious, and expensive composition — the flagship offering that pushes capability frontiers at premium cost. Alongside [[Claude Sonnet]] (medium-length, structured, reliable) and [[Claude Haiku]] (short, dense, cheap), the trio was introduced with the Claude 3 generation in March 2024. ## Current version: Opus 4.8 Released May 28, 2026. Key specifications: 1M-token context window (~555,000 words after tokenizer adjustments), adaptive thinking (extended reasoning built into the model's inference process), and pricing at $5/$25 per million input/output tokens — 5x the cost of [[Claude Sonnet]] and 25x [[Claude Haiku]]. Opus 4.8 is the fallback model when [[Claude Fable 5]]'s safety guardrails trigger on sensitive queries (fewer than 5% of sessions). ## Release timeline | Version | Date | Notes | |---------|------|-------| | Claude 3 Opus | March 2024 | First Opus-tier model; established the three-tier naming system | | Claude Opus 4 | May 2025 | Generation 4 launch alongside Sonnet 4 | | Claude Opus 4.6 | February 2026 | | | Claude Opus 4.7 | April 2026 | | | Claude Opus 4.8 | May 2026 | Current version; serves as Fable 5 safety fallback | ## KB relevance Opus matters for the vault in two ways. First, it was the model tier at the center of the [[Anthropic Government Ban]]: when Defense Secretary [[Pete Hegseth]] demanded Anthropic remove military-use restrictions from its models in February 2026, it was Opus-class capabilities the Pentagon wanted unrestricted access to for [[Project Maven]] and other defense AI programs. Anthropic's refusal to comply — and the government-wide ban that followed — is the foundational conflict in the KB's AI governance thread. Second, Opus 4.8's role as the safety fallback for [[Claude Fable 5]] means it is the model that actually serves users when Fable 5's guardrails trigger. When the [[Anthropic Model Export Directive]] suspended Fable 5 for all customers, Opus 4.8 effectively became the most capable Claude model available to anyone — the ceiling reverted from Mythos-class to Opus-class overnight. ## See Also - [[Claude Sonnet]] — the balanced mid-tier model - [[Claude Haiku]] — the fast, cost-effective model - [[Claude Fable 5]] — the Mythos-class model that uses Opus as its safety fallback - [[Claude Mythos 5]] — the restricted cybersecurity variant - [[Anthropic]] — the company behind the model family - [[Anthropic Government Ban]] — the Pentagon contract ban over Opus-tier military-use restrictions