Updated
Updated · Anthropic · Jul 23
Anthropic Launches Claude Opus 5 at $5 per Million Tokens, Nearing Fable 5 at Half Cost
Updated
Updated · Anthropic · Jul 23

Anthropic Launches Claude Opus 5 at $5 per Million Tokens, Nearing Fable 5 at Half Cost

3 articles · Updated · Anthropic · Jul 23

Summary

  • Claude Opus 5 is available now across Anthropic platforms, becoming the default model on Claude Max and the top model on Claude Pro.
  • At $5 per million input tokens and $25 per million output tokens—the same as Opus 4.8—Anthropic says it delivers frontier-level coding and knowledge-work performance at roughly half Fable 5’s cost.
  • Benchmarks cited by Anthropic show Opus 5 more than doubling Opus 4.8 on Frontier-Bench, scoring three times the next-best model on ARC-AGI 3, and matching or beating rivals at lower cost on tasks from automation to computer use.
  • Anthropic also framed Opus 5 as its most aligned model yet, with a 2.3 misalignment score, while saying it still trails Mythos 5 on exploit development and advanced biology work.
  • The launch also adds beta API features including mid-conversation tool changes and automatic fallbacks, with flagged Opus 5 requests able to route to Opus 4.8 by default.

Insights

Anthropic claims Opus 5 matches elite AI intelligence at half the cost, but what hidden workflow trade-offs are users actually paying for?
If Opus 5 deliberately trails Mythos 5 in cybersecurity, are specialized AI models becoming the new gatekeepers for high-risk digital threats?
With new evaluator loops and mid-task tool switching, is Opus 5 quietly transitioning Claude from a chatbot into an autonomous digital employee?