5 things you should know about Anthropic’s Claude Opus 5

1. Anthropic has launched Claude Opus 5, which the startup says comes close to Claude Fable 5’s performance at half the price. In some benchmarks, Opus 5 even outperforms Fable 5.

2. Unlike Fable 5 and Mythos 5, Opus 5 is designed for everyday use, meaning it works more efficiently and so Anthropic would rather you use it if you’re not doing specialized work, such as in cybersecurity. Opus 5 is the new default model on Claude Max, and the strongest model available on Claude Pro.

3. Anthropic expects that for Opus 5, the classifiers will “intervene around 85% less often than they do for Fable 5.” Like previous Opus models, Opus 5 does not have data retention requirements for general access.

4. Opus 5 improves coding, reasoning efficiency, and prompt-cache-friendly tool use. Those improvements won’t cost you more, as Anthropic has priced the latest model at $5 per 1 million input tokens and $25 per 1 million output tokens, just like Claude Opus 4.8.

5. Anthropic says Opus 5 is the startup’s “most aligned model to date.” It’s also Anthropic's fourth model release in less than two months, showing just how rapidly the startup is pushing the envelope.

Anthropic is likely accelerating its release schedule because it wants to show dominance in the space and revenue momentum, especially ahead of its IPO. Even if the startup doesn’t go public this year, those will still come in handy if it needs to raise yet another funding round.

I originally published this post as a video on LinkedIn and YouTube.