Introducing Claude Opus 5
For anyone doing complex, high-stakes knowledge work — detailed reports, technical problem-solving, thorough research — Opus 5 is now the most capable model available, nearly matching the top-tier Fable 5 at about half the price, and it is the default on Claude Max and the strongest option on Pro. If you have been reaching for a cheaper model to save cost, this narrows the gap enough that it is worth trying Opus 5 for your hardest tasks.
What changed
Source summary
Anthropic released Claude Opus 5, a new flagship model described as thoughtful and proactive, approaching the frontier intelligence of Claude Fable 5 at roughly half the price. It is state-of-the-art on coding and knowledge-work benchmarks (Frontier-Bench, GDPval-AA), priced the same as Opus 4.8, and is now the default model on Claude Max and the strongest available on Claude Pro.
Why it matters to your work
Practical interpretation
For anyone doing complex, high-stakes knowledge work — detailed reports, technical problem-solving, thorough research — Opus 5 is now the most capable model available, nearly matching the top-tier Fable 5 at about half the price, and it is the default on Claude Max and the strongest option on Pro. If you have been reaching for a cheaper model to save cost, this narrows the gap enough that it is worth trying Opus 5 for your hardest tasks.
When this could help
Useful as a default choice for coding, agentic or knowledge-work tasks where you want close-to-frontier quality without Fable 5's price, or via the API when you need a cost-effective high-capability model.
Claude Opus 5 is now Anthropic's default model on Claude Max and its strongest model available on Claude Pro, described by Anthropic as coming close to Claude Fable 5's frontier intelligence at half the price. Anthropic reports strong coding results (surpassing other models on Frontier-Bench v0.1, within 0.5% of Fable 5's peak CursorBench score at half the cost) and gains across knowledge-work, agentic and science evaluations, plus the lowest deception rate of any Anthropic model in its automated behavioural audit.
How to get started
On Claude Max, Opus 5 is used automatically as the default model; on Claude Pro, select it as the strongest available model.
Provider guidance ↗Via the API, call claude-opus-5 directly, or use Fast mode when you need faster responses and are willing to pay roughly double the base price for it.
Provider guidance ↗
Try this · The SCALA method
Give AI a clearer brief
- Situation
- You have a coding, research, or knowledge-work task and want strong model quality without paying for the top-tier model.
- Consultant
- Act as a careful technical assistant.
- Action
- Work through the task I give you, check and iterate on your own output before presenting it as final, and flag explicitly if you're uncertain about any part of the result.
- Language
- Direct, technically precise, with your own verification step shown where relevant.
- Ask
- Ask me for any missing context or constraints before starting.
SCALA suggested prompt · Adapt the context and check the result.
Learn the SCALA method →What to check
- Benchmark results (Frontier-Bench, CursorBench, GDPval-AA, ARC-AGI 3, Zapier AutomationBench, OSWorld 2.0, life-sciences evaluations) are from Anthropic's own internal runs, not independently verified third-party scores. Source ↗
- Cybersecurity safety classifiers on Opus 5 are less restrictive than on Fable 5; flagged requests in Claude.ai, Claude Code and Claude Cowork fall back to Opus 4.8 by default. Biology requests blocked on Fable 5 now route to Opus 5 rather than Opus 4.8, with stated limitations on long-running autonomous research tasks. Anthropic says Opus 5 remains behind its Mythos 5 model specifically in biology research and offensive cybersecurity. Source ↗
Sources & provider guides
Official documentation checked 12 Oct 2026. SCALA prompts and workflows are suggestions, not claims of hands-on testing.
Source & availability
This is a summary of a published provider announcement, with practical context from SCALA’s feed.
Plan, region, workspace and rollout conditions may affect access in Australia. Check the original source for current availability.
www.anthropic.com