Claude Opus 5 lands at Opus 4.8 pricing and becomes the Max default
Anthropic has released Claude Opus 5. The company says it approaches Fable 5’s intelligence at half the price and posts industry-leading scores on the Frontier-Bench and GDPval-AA benchmarks for coding and knowledge work. On Frontier-Bench v0.1, Opus 5 beats Opus 4.8 by more than 2x at a lower cost per task.
Pricing matches Opus 4.8: $5 per million input tokens and $25 per million output tokens. A Fast mode runs about 2.5x faster at double the price. Opus 5 is now the default on Claude Max and the strongest model on Claude Pro.
In automated behavioral auditing, Opus 5 scored 2.3 on undesirable behavior, the lowest of any recent model, though it still trails Mythos 5 on cybersecurity. It’s available today across all platforms as claude-opus-5.
Up next
csm moves Claude Code sessions between machines
Claude Code pins your session to local disk, so starting on a MacBook and moving to an Ubuntu server loses the context. The open-source csm tool makes sessions portable via a leader server over WebSocket, with backups to any S3-compatible store. MIT-licensed.
Related stories
- Anthropic: delete the “double-check your answer” lines from your Opus 5 prompts
- Opus 5 ships with a 200K default context window, down 80%
- Opus 5 tops Fable 5 on OSWorld 2.0 at a third of the price
- Fast mode is Claude API only, and limited to three Opus models
- A Claude Opus 5 model string showed up in Google’s Vertex catalog
- Qwen 3.8 and Claude Opus 5 show why raw benchmark scores don’t predict the bill
Comments
No comments yet. Be the first.
Join the conversation
Sign in to leave a comment. It appears after moderation.