Anthropic Halves Flagship Pricing With Opus 5
Same subscription, sharper model. If you're already paying for Claude, this upgrade landed in your account for free.
Founder, Automation Squad ·
The five-minute Opus 5 check
Opus 5 may already be sitting in a plan you pay for. This finds out, and tells you whether it is actually better for your work.
Open Claude and look at the model picker
Top-left in Claude.ai. Opus 5 is the default on Claude Max and the strongest option on Claude Pro — but only if it is the one selected.
Switch to Opus if it is not already selected
No upgrade, no new subscription. If you pay for Max or Pro, this is included at the plan you are already on.
Run the critique prompt below on a real document
Use something with actual stakes — a memo, a forecast, a board paper. A toy example will not tell you anything useful.
Compare the answer to your previous model
Same document, same prompt. If the answer is not visibly sharper, keep whatever you were using and revisit next release.
Take it with you
I'm going to give you [a real document, memo, or dataset from my business]. Read it critically and tell me: 1) the single weakest assumption in it, 2) one number I should double-check before I trust it, 3) what a skeptical board member would ask about it.
Anthropic released Claude Opus 5 on July 24, 2026. It's now the default model on Claude Max and the strongest option on Claude Pro — no action needed if you already pay for either plan.
The facts: Opus 5 costs $5 per million input tokens and $25 per million output tokens — the same rate as the outgoing Opus 4.8, and half what Anthropic charges for its actual flagship, Fable 5. It's available today across Claude.ai, the Claude API, Claude Code, and Claude Cowork. It's Anthropic's fourth Claude 5-series release in under two months. Anthropic says it runs at roughly 2.5x the default speed in Fast mode.
Automation Squad's take: model launches used to be quarterly events with keynote energy. Now they're closer to a rolling software update, and pricing is moving down, not up. The practical upshot for an exec is boring but real — a materially better model may already be sitting in a plan you're paying for, unused, because nobody told you to go look. Worth five minutes to check before you renew, upgrade, or shop a competitor.
Run this now: open Claude, check which model is selected (top-left picker in Claude.ai), and if it's not Opus, switch to it. Then paste this into a fresh chat: "I'm going to give you [a real document, memo, or dataset from my business]. Read it critically and tell me: 1) the single weakest assumption in it, 2) one number I should double-check before I trust it, 3) what a skeptical board member would ask about it." Compare the answer quality to whatever model you were using before.
Sources
Further reading
- Claude pricing and plans — Confirm which plan you're on and whether Opus is included before you build a workflow around it.