Opus 5: Cheaper AI coding power raises new security and budgeting questions

TL;DR Summary
Anthropic's Opus 5 delivers state-of-the-art coding performance for just $5 per million input tokens and $25 per million output tokens, outperforming Fable 5 on benchmarks and enabling longer autonomous tasks. However, the lower running cost intensifies concerns about security, spend management, and oversight for agentic AI, prompting a shift toward short-lived credentials, microVM isolation, and smarter telemetry; operators may route unsafe tasks to Opus 4.8 via Automatic Fallbacks, while balancing biology tasks and long-horizon workflows.
- Opus 5 costs a third of the price — and that’s actually the problem The New Stack
- Introducing Claude Opus 5 Anthropic
- Anthropic's new AI model rivals Fable 5 and is cheaper as businesses fret about costs CNBC
- Claude Opus 5 is now available in GitHub Copilot The GitHub Blog
- Anthropic’s Opus 5 is about token efficiency, not a capability leap Ars Technica
Reading Insights
Total Reads
0
Unique Readers
0
Time Saved
72 min
vs 73 min read
Condensed
99%
14,532 → 75 words
Want the full story? Read the original article
Read on The New Stack