What Is Claude Opus 5? Anthropic's Everyday Flagship Explained
Anthropic didn't pitch Claude Opus 5 as its biggest or most expensive model ever. It pitched it as the model built to be used every day — close to the intelligence of Claude Fable 5, at half the price. This guide covers what actually changed from Opus 4.8, what it costs, where it's available, and who should (and shouldn't) bother switching.
Opus 5 is now the default model on Claude Max and the strongest model available on Claude Pro.
Claude Opus 5 Is the New Everyday Opus Model
Claude Opus 5 is Anthropic's latest model in the Opus tier — the line built for complex reasoning, coding, and agent work. It replaced Opus 4.8 as the flagship on July 24, 2026, and Anthropic's own framing was unusually specific: not "our smartest model ever," but a model that gets close to Fable 5's intelligence at roughly half the cost, designed to be run every single day rather than saved for the hardest tasks.
That distinction matters more than it sounds. Fable 5 sits in Anthropic's separate Mythos tier, priced and restricted differently, and most people never touch it directly. Opus 5 is the model actually replacing what most Claude Pro and Claude Max subscribers were already using — so for the vast majority of people asking "what's new in Claude," Opus 5 is the real answer.
What Actually Changed From Opus 4.8
Adaptive thinking, on by default
Opus 5 automatically scales how much it reasons based on how hard the question is. Simple requests get fast answers; multi-step problems get visibly deeper reasoning before it commits to a response, without you having to ask for it.
A manual effort dial
You can also override that and set effort to low, medium, or high yourself — trading speed and cost for deeper reasoning on tasks where you know you need it, or keeping things fast and cheap when you don't.
Better self-correction on agent tasks
Anthropic says the model verifies its own work and recovers from errors mid-task more often without a human stepping in — relevant if you're running Opus 5 inside Claude Code or as part of a longer agent workflow.
Same price as Opus 4.8, meaningfully more capability
Anthropic didn't raise Opus pricing for this jump. On Frontier-Bench, the company says Opus 5 more than doubles Opus 4.8's score at a lower effective cost per completed task.
Opus 5 vs. Fable 5 vs. Sonnet 5
Claude Fable 5 (Mythos tier)
Anthropic's most capable public model. Opus 5 comes close to it on most benchmarks and actually beats it on some — but Fable 5 still leads specifically on cybersecurity-related tasks, where its safety tuning goes further.
Claude Opus 5 (this model)
The new default on Claude Max, and the strongest model on Claude Pro. Built for daily use across coding, knowledge work, and agent tasks at a price Anthropic didn't increase from the prior Opus.
Claude Sonnet 5
Cheaper and faster, still strong on everyday coding, tool use, and reasoning. For short drafts, simple classification, or boilerplate code, Sonnet 5 often performs close enough to Opus 5 that the price gap isn't worth it.
Claude Haiku 4.5
The fastest, lightest option — still waiting on its own "5" upgrade as of this writing. Best for high-volume, low-complexity tasks where speed matters more than depth.
Where Opus 5 Actually Struggles
Benchmark charts told one story at launch; the first week of real-world use told a slightly different one. Multiple early reviewers, including the team at Every, described Opus 5 as capable but occasionally frustrating in practice — stopping short of finishing a task, getting unusually cautious about edits it wasn't sure about, or needing more back-and-forth than Opus 4.8 did on certain workflows.
None of that erased the underlying gains — most reviewers who tested it against six or seven competing models still ranked it near or at the top on real coding and agent benchmarks. But it's a reminder that a model that wins on paper doesn't always feel effortless on day one, especially if your existing prompts and workflows were tuned around the previous model's quirks.
How to Get Claude Opus 5
Claude Max
Opus 5 is the new default model — you're already using it if you haven't changed your model setting.
Claude Pro
Opus 5 is available as the strongest model on the plan; you may need to select it manually depending on your current default.
Claude API / Platform
Available now as claude-opus-5 at $5 per million input tokens and $25 per million output tokens — identical list pricing to Opus 4.8.
Bottom Line
Claude Opus 5 is Anthropic's new everyday flagship: it comes close to Fable 5's intelligence at half the price, with no increase over Opus 4.8's rates, and an adjustable effort dial that lets you trade speed for depth as needed. It's the strongest model most Claude Pro and Claude Max users will actually use day to day — but it isn't automatically the right tool for simple tasks, where Sonnet 5 or Haiku 4.5 will usually get the job done for less.
More AI Guides, Written Clearly
Browse our full library of AI tool guides, updated regularly as models and features change.
Common Questions
What is Claude Opus 5?
Claude Opus 5 is Anthropic's current Opus-tier model, released July 24, 2026. It's built to approach the intelligence of Claude Fable 5 at roughly half the price, and it's now the default model on Claude Max and the strongest model available on Claude Pro.
How much does Claude Opus 5 cost?
On the API, Claude Opus 5 costs $5 per million input tokens and $25 per million output tokens — the same list price as its predecessor, Opus 4.8. It's included in Claude Pro and Claude Max subscriptions.
Is Claude Opus 5 better than Claude Fable 5?
Not quite, but it gets close on many benchmarks at half the cost. Fable 5 remains ahead on cybersecurity tasks specifically, and both trail Claude Mythos 5, which isn't publicly available.
What is the "effort" setting in Claude Opus 5?
It's a toggle — low, medium, or high — that lets you choose how much reasoning Opus 5 spends on a task. Lower effort is faster and cheaper; higher effort trades speed and cost for deeper reasoning on harder problems.
More from the creator
Explore two other projects from the same creator.