Claude Opus 4.8: Faster, Smarter, and Same Price

Summary

Anthropic has launched Claude Opus 4.8, promising enhanced judgment, increased honesty about its progress, and longer independent work capability, all at the same price as its predecessor. This update effectively reduces cost per performance, especially with the introduction of a new fast mode that is roughly two and a half times faster, now also three times cheaper than before. In coding benchmarks like Swebench Pro, Opus 4.8 saw a significant five-point jump, reaching sixty-nine point two percent. However, on terminal navigation benchmarks, GPT 5.5 still holds the lead. A major new feature is 'dynamic workflows' in Claude Code, designed for complex tasks. This allows Claude to create plans, run hundreds of parallel sub-agents, and verify its work before reporting, potentially handling tasks like large-scale code migrations or bug hunts end-to-end. This capability is enabled by Anthropic's increased compute access through deals with XAI and cloud providers. Pricing for standard usage remains unchanged at five dollars per million input tokens and twenty-five dollars per million output tokens, with fast mode now costing ten dollars per million input tokens and fifty dollars per million output tokens.