Opus 5 pushes Anthropic into a price war for the everyday model

Anthropic released Claude Opus 5 after several weeks of industry tension, including its latest clash with the US government and a recent OpenAI security incident, according to The Verge. The company says Opus 5 approaches Claude Fable 5 capabilities at half the price. API pricing is $5 per million input tokens and $25 per million output tokens.

Anthropic is turning Opus 5 into a working default, not a display model

The placement matters. Anthropic says Opus 5 is the new default for Claude Max and the strongest model in Claude Pro. It still describes Fable 5 as the option for the most ambitious work and long-horizon agent projects.

That is the product message: the best intelligence cannot live only as an expensive capability demo. Anthropic wants more intelligence in the layer people use every day: coding, analysis, knowledge work and business automation.

Token price is turning into cost per completed task

Opus 5 costs the same as Opus 4.8 and The Verge reports that it is slightly cheaper than OpenAI GPT-5.6. Fast mode, now in research preview, offers higher speed at double the base price.

For buyers, this is not just a token table. If the model handles more steps without escalating to the more expensive Fable 5, it can lower the cost of agentic work. If it needs more reviews, fallbacks and manual repair, cheaper tokens lose their charm quickly.

Safety messaging has to reassure customers and regulators

Anthropic emphasizes stronger cyber safeguards than Opus 4.8, calling Opus 5 its most aligned Opus model and the least susceptible to being tricked into misuse. The company also says it continues to work with government partners on independent testing.

That is practical, but it also shows the tension around this generation of models. The more useful a model becomes for long agentic tasks, the more customers need audit, permissions and clear fallback rules.

The market will test whether Max and Pro can carry frontier intelligence

The next signal is not only a benchmark. Watch whether Claude Max and Pro users feel a higher work ceiling without suspecting hidden usage caps or facing sharp price changes. If Opus 5 holds as a daily model, Anthropic gains leverage against competitors. If not, Fable remains an expensive poster on the wall.

Lilith's verdict

Opus 5 is an attempt to put expensive intelligence on an ordinary shift without handing finance a terrifying receipt. The accountant will not read the launch post. They will count closed tickets.